D&D 5e 2014 Creating Adventures Guide: Step-by-Step Design

Introduction

Chapter 3 of the 2014 Dungeon Master's Guide treats adventure design the way a TV writers' room treats an episode: a single adventure has its own beginning, middle, and end, but it also has to connect to the larger campaign it belongs to. This guide walks through that chapter in order - the qualities every strong adventure shares, how to structure one from hook to climax, the difference between building around a location and building around a villain's plan, and how to turn all of that into individual encounters your players will actually remember.

None of this requires drawing a single map before you understand what you're building. The chapter's own steps work whether you're planning a dungeon crawl for Saturday or sketching the outline of something that will take your table six sessions to finish.

Elements of a Great Adventure

The DMG opens with a short list of qualities that show up in adventures players remember - not a checklist to hit every time, but a set of levers worth pulling when an adventure feels flat.

A credible threat. Something worth the heroes' attention: a single villain, a villain with lackeys, a whole evil organization, or an assortment of monsters. Whatever form it takes, give the threat goals the party can actually uncover and thwart, not just a name on a stat block.

Familiar tropes with a twist. Dragons, orcs, and a wizard in a tower are staples for a reason, and starting an adventure in a tavern is a running joke because it works. The DMG's advice isn't to avoid the familiar - it's to put an occasional spin on it. The mysterious cloaked figure offering a quest on the king's behalf might actually be the king in disguise. The "insane wizard" terrorizing a tower might be an illusion a band of gnome thieves built to guard their loot.

A clear focus on the present. History can set an adventure in motion and reward the party for digging into it, but the adventure itself should stay focused on the current situation: what the villain is doing right now, and how the party gets pulled into stopping it. Let the world's past show up as texture, not as the main event.

Heroes who matter. The party's decisions need to be able to change the outcome. An adventure that resembles a novel or TV episode can still railroad players if it only allows one ending - if a major villain shows up before the climax, the story needs to actually permit the party defeating them right there, not just funnel everyone toward a scripted final battle.

Something for every player type. Balance the three basic activities - exploration, social interaction, and combat - and most tables find something to enjoy. This doesn't mean forcing every activity into every adventure regardless of who's playing: an adventure built for a table with zero interest in combat doesn't need a maximum amount of fighting just to satisfy an abstract player-type checklist.

Surprises. A ruined castle on a hill that turns out to hide a dragon's tomb underneath. A wilderness trek that leads to a tower that only appears on full-moon nights. Players remember locations and moments like these. The DMG's caution runs the other way too - too many surprises stop feeling surprising, so save the occasional twist (goblin sappers with kegs of oil strapped to their backs, an unexpected guest crashing the villain's estate) rather than stacking them into every scene. Sometimes a straightforward fight with an orc guard is exactly the encounter a scene needs.

Useful maps. A dungeon with branching corridors and real decision points gives players agency; a wilderness map dotted with landmarks beats an unbroken expanse of blank terrain. If drawing isn't your strength, borrowed maps, real-world floor plans, and mapmaking software all work as a starting point.

Published Adventures

Not every adventure needs to be built from scratch. A published adventure hands you a pregenerated scenario - maps, NPCs, monsters, and treasure already assembled - and the D&D Starter Set is the most common example a new table encounters first.

Published material doesn't have to be run exactly as written. Swap in a villain your players have already crossed paths with, graft a detail from your own campaign onto the background so the adventure feels like it belongs to your table specifically, or use it purely as inspiration - repurposing a temple map with your own choice of monsters, or borrowing a chase sequence as a template for a pursuit scene of your own. Because a published adventure can't anticipate every action your specific party takes, using one often frees up your prep time to focus on the plot threads unique to your campaign that the module itself can't address.

Adventure Structure

Like most stories, a typical adventure breaks into a beginning, a middle, and an end.

Beginning. An adventure starts with a hook - a compelling, specific reason for the characters to get involved. The party stumbles onto something they weren't meant to see, monsters ambush them on the road, an assassin makes an attempt on their lives, a dragon lands at the city gates. A strong opening is exciting and focused enough that players leave the session already looking forward to the next one, with a clear sense of where the story is headed.

Middle. This is where most of the adventure's actual content lives. Each new challenge should hand the characters a choice with a clear effect on how the story ends. Characters might discover a rumor of treasure was actually a trap, or that the "spy" in the queen's court is really the queen's own scheme to consolidate power - revelations that change the party's understanding of the situation, not just add flavor text. While the party works to stop the villain, the villain should be working just as hard to carry out their own plan, whether that means hiding evidence, misleading the party, or confronting problems head-on. The one thing to avoid: letting the characters become spectators to events they can't actually influence.

Ending. The climax is the scene where all the built-up tension peaks - the characters' fate, and often much more, hanging in the balance. The outcome should hinge on the party's actual choices, never feel like a foregone conclusion. And an ending doesn't need to resolve everything: leaving a thread or two hanging is an easy, natural bridge into whatever adventure comes next.

Location-Based Adventures

An adventure can be built around a location or around an unfolding sequence of events. Location-based adventures - crumbling dungeons, remote wilderness sites, fortified settlements - are the backbone of most classic D&D campaigns, and building one breaks down into a repeatable set of steps you can tackle in any order.

1. Identify the party's goals. What draws the characters into this specific dungeon, wilderness area, or settlement? A goal can be as direct as recovering a stolen item or as open-ended as discovering why a villain cares about the location in the first place. Matching a concrete goal to the location gives players a reason to engage with it beyond "the DM put a dungeon here."

2. Identify the important NPCs. Villains, allies, and patrons each need enough definition to feel real at the table - not full backstories for every name, just enough that the party's interactions with them land. A villain's motivation matters more than their stat block; an ally's usefulness matters more than a full character sheet if they're never going to fight. Villains span a wide range: a beast or monstrosity with no real agenda beyond survival, an aberration bent on corruption, a dragon after domination and plunder, or a humanoid schemer working an angle from inside the party's own trusted circle. Allies range from a skilled fellow adventurer to an enthusiastic but inexperienced commoner, and patrons range from a retired adventurer with old contacts to a desperate merchant with more coin than sense. Chapter 4's guidance on creating NPCs is the natural next stop for fleshing any of these out beyond a name and a role.

3. Flesh out the location details. A dungeon, wilderness area, or settlement all need enough concrete detail to run - notable features, inhabitants, hazards, points of interest - without demanding a fully mapped-out world before the first session can happen.

4. Find the ideal introduction. The best hooks arise naturally from the goal and the setting rather than feeling bolted on. A social encounter where the party learns what they need to do and why, a surprise attack that drags them in, or stumbling onto information by accident can all work. The DMG's own examples range from mundane (a stranger in a tavern points the party toward the location) to outright supernatural (the party shares the same dream about the location three nights running, or a ghost appears and can only be put to rest by someone entering it). The throughline is that the hook should feel like it grew out of the situation, not like a DM announcement that the adventure is starting now.

5. Consider the ideal climax. The climax should fulfill the promise of everything that came before it, hinging on the party's actual successes and failures up to that point rather than an outcome decided in advance. A few shapes come up often enough to be worth stealing outright: a straight battle against the villain and a group of minions, a chase that ends in a final confrontation at the villain's refuge, a race to reach the villain before their plan completes, or a dungeon that starts collapsing while the fight against the main villain is still underway. Whichever shape fits, the DMG's caution about betrayal climaxes applies broadly - a trusted ally turning on the party at the worst possible moment is a strong beat used once, and a cheap trick used twice.

6. Plan encounters. With the location and overall story in place, key individual encounters to specific areas on the map. Each room or wilderness zone gets its own description - physical features, and whatever encounter (if any) plays out there. This is what turns a plain sketch of numbered rooms on graph paper into a real adventure. See Creating Encounters below for how to build the individual scenes.

Hot Take: Event-based structure gets treated like the more sophisticated option because it reads better on paper - a villain's timeline, a flowchart of contingencies - but for most tables, especially anyone still building confidence behind the screen, location-based design is the stronger default. A map absorbs player unpredictability by design: wherever the party wanders, they wander into content, because the content is pinned to the ground instead of to a schedule only you know. Event-based adventures ask you to improvise a villain's reaction to player choices in real time, which is a harder skill to run well under pressure. Start with a map. Move to timelines once you've got a session or two of confident improvisation behind you.

Event-Based Adventures

In an event-based adventure, the focus shifts from where things happen to what the characters and the villain actually do - location becomes secondary. Building one takes more work than a location-based adventure, but it breaks down into its own set of steps.

1. Start with a villain. The villain drives an event-based adventure more directly than in a location-based one, so investing effort here pays off. An undead villain seeking revenge for a centuries-old injury, for instance, opens up motivations a straightforward monster raid never would.

2. Determine the villain's actions. Build a timeline of what the villain does and when, assuming no interference from the party. The DMG groups villain action patterns into a few recognizable shapes: a big event the villain is building toward (a festival, a coronation, an eclipse), a crime spree that escalates in boldness over time, growing corruption that spreads wider as time passes, a one-and-done crime followed by an attempt to avoid consequences, serial crimes repeated in a consistent pattern rather than escalating, and a step-by-step plan carried out as a specific sequence of actions. Picking one of these shapes early makes the villain's behavior easier to run consistently once the party starts interfering.

The book's own worked example is worth walking through because it shows how much a single villain concept can carry. Start with a vampire imprisoned centuries ago by a religious order of knights, now free and seeking revenge against the order's current members. That's a step-by-step villain: the vampire slips past locked doors in gaseous form, murders knights one at a time, and makes the deaths look natural at first. As a twist on top of the basic revenge plot, the vampire's real goal is retrieving a necklace set with nine gems from among its stolen jewels - gems that can open a gate to the Nine Hells, which means "defeat the vampire" and "stop the vampire's actual plan" turn out to be two different problems layered on top of each other.

3. Determine the party's goals. What does the party actually need to do about the villain's plan - bring them to justice, protect an object or NPC, uncover a conspiracy, negotiate a peace? A goal here can also suggest exactly how the party gets pulled into the villain's plans in the first place.

4. Identify important NPCs. Event-based adventures typically need a deeper NPC roster than location-based ones, since many characters won't fall neatly into "ally" or "patron" - their attitude toward the party stays undecided until the party actually interacts with them.

5. Anticipate the villain's reactions. As the party pursues their goal and interferes with the villain's plans, how does the villain respond - lash out, send a warning, improvise a workaround? Walk through the timeline from step 2 and think through the villain's likely response to each way the party might disrupt it. A flowchart branching off the timeline (or a separate one tracking villain reactions to specific interferences) is a practical way to keep this straight at the table.

6. Detail key locations. Since the location isn't the star of an event-based adventure, these can be smaller and simpler than a full dungeon complex - specific rooms or sites where combat or major exploration will actually happen, like a throne room, a guild headquarters, or a crumbling manor.

7. Choose an introduction and a climax. The same Adventure Introduction and Adventure Climax concepts that anchor a location-based adventure's opening and ending work just as well here.

8. Plan encounters. In an event-based adventure, encounters happen wherever the villain's agenda crosses the party's path rather than at fixed map locations. You can't always predict exactly when or where that intersection will occur, but you can prepare a list of likely encounters - descriptions of the villain's forces, their lieutenants and minions, and scenes tied to the key locations from step 6.

Mysteries

A mystery is an event-based adventure built around the party investigating a crime, usually a robbery or murder. Unlike a mystery novelist, you can't fully predict what your players will do with the clues you hand them, so build in redundancy rather than betting everything on one clean deduction chain.

Villains following a crime-spree, one-and-done, or serial-crimes pattern naturally suggest a mystery. Building one adds three elements on top of the standard event-based steps:

The victim. A victim connected to the villain in some meaningful way makes for a stronger mystery than a random killing, since part of the fun is uncovering the twisted relationships that led to the crime. A victim connected to one of the player characters is an especially reliable way to pull the party in - sometimes even making them suspects themselves.

The suspects. Populate the mystery with NPCs who didn't commit the crime but plausibly could have - motive, means, or opportunity. A "closed circle" of suspects (a finite, defined group who are the only people who could have done it) is a classic structural choice borrowed straight from detective fiction. Giving more than one innocent suspect a secret of their own keeps players guessing, since an innocent person acting nervous or evasive under questioning reads exactly like guilt.

The clues. Clues connect a suspect to the crime by pointing at motive, means, or opportunity - verbal clues from witness and suspect statements, physical clues like an unfinished message or a dropped weapon. Some clues should deliberately point at the wrong suspect, forcing the party to keep digging rather than accepting the first plausible answer. Err toward too many clues rather than too few: a party that solves the mystery faster than expected still feels satisfied, while a party that can't find enough clues gets frustrated instead.

Intrigue

Intrigue adventures are event-based adventures centered on power struggles - common in royal courts, but just as workable in a merchants' guild, a crime syndicate, or a temple hierarchy. Rather than dark plots and monster threats, intrigue runs on the exchange of favors, the rise and fall of influence, and the honeyed language of diplomacy.

An intrigue adventure only works if the party is actually invested in the stakes. A struggle over who becomes the king's chamberlain falls flat if the players don't care who the king's chamberlain is - but the same struggle lands if winning the chamberlain's favor means the party can call on royal soldiers to defend their own stronghold later.

Building an intrigue adventure follows the same event-based steps with two real differences. Villains can range from a single schemer (a noble plotting a monarch's assassination) to a whole cast of rival villains, each independently pursuing power - or no villain at all, in adventures that run purely on the exchange of favors, in which case you skip straight to the party's goals and spend your prep time on the supporting cast instead. Influence becomes a resource worth tracking explicitly, whether you treat it like inspiration (earned through favors and demonstrations of power, spent for a mechanical benefit) or like renown within a specific faction.

Framing Events

An entire adventure can hang off a single framing event - a coronation, a solar eclipse, a festival, a trial, a wedding - that gives the party a concrete reason to be somewhere at a specific time and gives the villain's plan a deadline to build toward. The DMG's own list of sample events ranges from the civic (a royal address, a council meeting, the signing of a treaty) to the cosmic (a planar conjunction, a lunar eclipse, the arrival of marching modrons) to the personal (a wedding, a funeral, the birth of a child) - pick whichever scale matches how big this adventure is supposed to feel. A framing event does double duty: it's a built-in hook, and it's a natural climax deadline, since whatever the villain is planning tends to come to a head exactly when the event does.

Complications and Twists

Moral Quandaries

A moral quandary forces a choice no amount of combat skill or spellcasting can resolve - a problem of conscience where every option costs something. The DMG groups quandaries into a few recurring shapes: an ally quandary (two allies whose help the party needs both refuse to cooperate with each other), a friend quandary (a character the party cares about makes an impossible demand), an honor quandary (victory requires breaking a personal oath or code), a rescue quandary (stopping the villain means letting innocents come to harm, or vice versa), and a respect quandary (two trusted allies give directly conflicting advice, and picking one costs the other's respect). None of these need a mechanical resolution - the payoff is entirely in what the party chooses and lives with afterward.

Twists

A twist complicates the party's path to their goal without derailing the adventure entirely: a time limit suddenly imposed, a rival group racing toward the same objective, false information the party has been acting on, an ally who turns out to be a traitor (a twist the DMG specifically cautions against overusing), or a magical compulsion - a geas spell, for instance - forcing the party to see a goal through whether they still want to or not. A well-placed twist makes players adjust their plan mid-scene instead of executing the plan they walked in with, which is exactly the kind of moment that makes a session memorable afterward.

Side quests round out an adventure's texture without competing with its primary goal - a rumored item somewhere in the area, information an NPC is willing to trade, a specific monster worth hunting down. A side quest that succeeds can feed back into the main goal without becoming a required detour.

Creating Encounters

Encounters are the individual scenes that make up an adventure's larger story. First and foremost, an encounter needs to be fun for the players to play through; second, it shouldn't be a burden for you to run. Beyond that, a well-built encounter usually has a clear objective and ties directly into the campaign's overarching story, building on what came before while setting up what's still ahead.

Every encounter resolves in one of three ways - the characters succeed, partly succeed, or fail - and each outcome needs real consequences, so players feel like both their wins and their losses actually mattered.

Character Objectives

A transparent objective keeps players engaged instead of leaving them confused about what an encounter actually wants from them. The DMG offers several reusable objective shapes you can build individual encounters - or, chained together, entire multi-encounter obstacles - around:

• Make Peace. Convince two opposing sides (or their leaders) to end a conflict, complicated by the party having enemies on one or both sides, or by a hidden third party instigating the fight.
• Protect an NPC or Object. Act as bodyguards, complicated by an NPC who's cursed, diseased, too young or old to fight, prone to panic, or apt to make risky decisions of their own - or by an object that's sentient, cursed, or awkward to transport.
• Retrieve an Object. Gain possession of something specific in the encounter area, ideally before combat ends, complicated by enemies who want the same object and will fight for it.
• Run a Gauntlet. Get through a dangerous area, where reaching the exit matters more than defeating everything inside - complicated by a time limit, a wrong-turn decision point, traps, or hazards.
• Sneak In. Move through the encounter area without alerting anyone to the party's presence, with real complications kicking in only if they're detected.
• Stop a Ritual. Fight through the villain's minions before the ritual can even be interrupted, complicated by a time limit if the ritual is close to completing - and by whatever the ritual does if it finishes anyway.
• Take Out a Single Target. Choose between fleeing a heavily guarded villain to fight another day or pushing through the minions to reach them, complicated when those minions turn out to be innocents under the villain's control rather than willing followers.

Combat Encounter Difficulty

The DMG sorts combat encounters into four difficulty tiers. An easy encounter barely taxes the party's resources - victory is close to guaranteed, at the cost of a few hit points. A medium encounter has one or two scary moments, but the party should come out the other side without a death, possibly spending some healing resources along the way. A hard encounter could go badly - weaker characters risk getting knocked out, and there's a real if small chance of a character dying. A deadly encounter could be lethal for one or more characters outright, and getting through it usually takes good tactics and some luck.

Turning those four labels into an actual number - how much monster XP adds up to a medium encounter for your specific party, and how a group of monsters counts for more than the sum of their individual XP values - is table math the DM Quick Reference guide covers in full, including the XP Thresholds by Character Level table, the Encounter Multipliers table, and the Adventuring Day XP table for estimating how far a party can push in one day. Worth knowing without flipping over there: a monster with a challenge rating well above the party's level deserves extra caution regardless of what the math says, since a single high-CR creature can sometimes one-shot a lower-level character or shrug off an entire spell list through immunities the party's XP-based math doesn't account for.

Difficulty also shifts with situation, not just numbers. Give the characters a drawback the enemy doesn't share - surprise, no cover, an environmental hazard hitting only them - and bump the encounter up a step. Give them a benefit the enemy lacks and it drops a step; benefits and drawbacks that both apply cancel out. Terrain that both sides can exploit (a frayed rope bridge, a pit, a chandelier worth swinging on), enemies dug into hard-to-reach positions that force ranged characters to reposition, and mixed groups of monsters working together all add tension a flat difficulty rating can't capture on its own.

Random Encounters

As characters explore a wilderness area or dungeon, the unexpected is bound to show up eventually - random encounters are the DMG's tool for delivering it, usually rolled from a table built around a specific location and theme.

Random encounters have a reputation as filler, but a well-built table earns its place at the session by doing real work: creating urgency (a party that fears wandering monsters stops dawdling and starts looking for a safe place to rest), establishing atmosphere (a table stocked with bats, wraiths, and zombies tells the party what kind of danger is coming), draining party resources to build tension before a bigger fight, occasionally providing help instead of harm, revealing details about the world that foreshadow what's ahead, and reinforcing the campaign's larger themes - a border-war campaign's tables might swap in bedraggled soldiers and refugee columns depending on whether the party is in friendly or hostile territory.

Not every DM finds them worth the table space, and that's a legitimate call: if random encounters keep interrupting your pacing without adding anything, the book's own advice is to just stop using them.

Triggering them. Reach for a random encounter check when players are drifting off-track and slowing the session down, when characters stop for a rest, during a long uneventful journey, or when the party draws attention to itself when staying quiet would have served them better. A common approach is checking once every hour of in-fiction time, once every four to eight hours, or once per rest period - whatever matches how active the surrounding area actually is. A typical trigger method rolls a d20 and treats an 18 or higher as an encounter, then rolls again on the location's specific table (rerolling any result that doesn't make sense for the circumstances). Not every creature crossing the party's path counts as a random encounter - harmless wildlife and background NPCs going about their day don't need a table entry.

Building your own table. Start with a location the party is likely to pass through, then list creatures, NPCs, events, or discoveries that plausibly belong there. Think through why each entry would actually be encountered - on patrol, hunting, searching for something specific - since a monster that's just standing around for no reason reads as artificial. Mix in a few non-monster entries too: a strange landmark, an overheard rumor, or an unusual discovery keeps a table from being purely a combat-trigger list. The DMG's own sample table uses a 2d12-style bell-curve roll so mid-table entries come up more often than the extremes, which is worth copying if you want some encounters to feel like the area's signature threat and others to feel rare.

Random doesn't mean unfair. A random encounter table shouldn't include anything capable of steamrolling a party that had no way to prepare for it - a single monster with a challenge rating at or below the party's level, or a group whose adjusted XP lands in the easy-to-hard range, are the DMG's own guidelines for what belongs on the table. Not every monster encounter needs to end in combat, either: a young dragon circling overhead or a reclusive giant wandering the hills can both be encounters the party talks, hides, or bargains their way through instead of fighting.

The DMG's own sample table, built for a sylvan forest, shows the range a good table can cover in one page: a displacer beast and a gnoll war band sit at the dangerous end, a lone giant owl or an ivy-covered statue of an elven deity sit at the harmless end, and a patch of glowing berries that grant temporary invisibility sits somewhere in between as neither threat nor reward, just a detail worth discovering. That spread - real danger, pure flavor, and the occasional useful find, all keyed to one specific location - is the actual target when building a table of your own, not a wall-to-wall list of monsters waiting to fight the party on sight.

Bottom Line

Adventure design comes down to a short chain of decisions: pick the qualities that make this specific adventure worth playing, decide whether it's built around a place or a plan, give it a real beginning, middle, and climax, then break the whole thing down into individual encounters built around a clear objective. Location-based and event-based structures both provide a repeatable set of steps rather than a rigid formula, and mysteries, intrigue, complications, and twists are tools to reach for when a straightforward version of either structure needs more texture. None of it requires running the session mechanics - once an adventure is built, the Running the Game guide covers actually resolving what happens at the table, and the DM Quick Reference has the encounter-math tables this guide deliberately left out. For stocking the specific location an adventure is built around - a dungeon, a settlement, a wilderness site - that's a DM's toolbox problem, not an adventure-structure one. Return to the Dungeon Master Guides hub for the full 2014 DM chapter map.