Game design

The Legend of Zelda and Zelda II's Dungeon Design

The first two Zelda games established bosses, items, keys, locked doors, and maze-like dungeons, but they also show why later games had to refine the formula.

The first game establishes the basic shape

The Legend of Zelda was built around mystery and surprise. It asks the player to wander away from the starting screen, explore a strange sprawling landscape, find eight pieces of the Triforce, and rescue Princess Zelda.

Those pieces are hidden in dungeons, and right from the beginning the series was split between overworld and underworld. The player explores underground mazes, fights a boss, and earns a new piece of the Triforce.

Those labyrinths established a lot of what would later define Zelda dungeons. Each maze has a boss, an item, a map, a compass, and a collection of small keys and locked doors. Most rooms are combat challenges, but the larger goal is navigation: find your way through an interconnected maze without getting lost.

The rooms are not arranged in a straight line. They are joined together, broken up by locks and obstacles, and full of reasons to search for keys and items. The recognizable template is already there. The roughness is in how those ideas work.

Shared keys make dungeons less self-contained

Small keys became a series staple: they can open any normal locked door, and they disappear after use. In most later Zelda games, those keys belong to the dungeon where they are found. In the first Zelda, a key can be carried from one dungeon to another.

That sounds flexible, but it creates messy outcomes. Sometimes the player has more keys than they need because earlier dungeons contain extras and some locked rooms are optional. A dungeon can become too easy if the player walks in with a handful of keys already in reserve.

The opposite problem is worse. A player can enter a dungeon without enough keys, especially if they have opened every door they found or failed to search previous dungeons thoroughly. Some dungeons do not contain enough keys to open every door inside them.

Level 6 shows the issue clearly. Enter with no spare keys, pick up one key, and two locked doors appear. One door only leads to a hint room. Choose that door and the real path is blocked, which means leaving to find or buy another key.

Buying keys is a messy patch, especially when they are expensive and rupees are scarce. Zelda dungeons work better as self-contained spaces, so restricting small keys to their own dungeons was a strong later correction.

Required items need to be required before the boss

The first game also lets players finish many dungeons without collecting the key item inside. In Level 1, it is easy to miss the bow and arrow completely, and the player may not realize the mistake until Level 6, where arrows are needed to kill the spider boss.

The game already contains the solution to this problem. In Level 4, water blocks the route to the boss room, so the player must find the stepladder elsewhere in the dungeon before advancing. Anyone who beats that dungeon necessarily found the item.

That logic makes sense: optional rewards can be missed, but items required to finish the adventure should not be quietly skippable. Let players overlook the magic rod, upgraded candle, power ring, spell book, or magical key if those are optional. Do not let them leave the relevant dungeon without the raft if the game later depends on it.

This rule was not fully established immediately. Even A Link to the Past lets players finish the Tower of Hera without the Moon Pearl. Link's Awakening is where the structure becomes more consistent: the dungeon item is usually required to reach the boss.

Bomb walls turn from secrets into unfair progress gates

Bombs create an even messier problem. One common criticism of the first Zelda is that it asks players to burn every bush and bomb every wall. On the overworld, that complaint is not entirely fair. Most hidden rooms are optional, even if their heart containers and rupees are extremely helpful.

Only one required bush and one required wall hide dungeon entrances, and both are hinted. The overworld can be defended as a place of mystery where secrets reward experimentation.

Inside the dungeons, the same idea becomes much harder to defend. In the first four dungeons, bombable walls mostly create shortcuts, bypass locked doors, or reveal optional treasure rooms. That is fine. They are secrets, not mandatory routes.

Later dungeons make invisible bomb walls part of the required path to the boss. Now the player may waste bombs, grind for more, and draw maps just to track which walls have already been tested.

Dungeon 5 at least teaches the concept. The player is locked in a room, given bombs, and can see from the map that there is a room to the left. The intended lesson is clear: from this point on, walls may need to be destroyed to advance.

That warning is not enough. In Dungeon 7, reaching the boss requires bombing four different walls, and one early hidden room is not even shown on the map. Combined with rooms where one random block must be pushed only after every enemy is dead, the design drifts from challenging into unfair.

The early dungeon formula is important but unrefined

The first Zelda deserves credit for laying the groundwork: maze-like spaces, bosses, items, maps, compasses, keys, locks, secrets, and the rhythm between overworld exploration and underground trials.

But many of its dungeon systems are unrefined. Keys are too global. Required items are too easy to miss. Bomb walls hide mandatory progression. Puzzles often mean pushing a random block after clearing a room. The later games did not abandon this foundation. They fixed it.

The overworld mystery remains powerful, but some of the late dungeons show why future entries needed clearer rules, fairer signaling, and more self-contained dungeon logic.

The second game changes almost everything

Zelda II is a strange sequel. It turns the adventure into a side-scrolling platformer with a dedicated jump button, while the overworld remains top-down and includes random battles. It adds precise sword fighting, magic spells, experience points, leveling, grinding, and brutal difficulty.

Almost everything about it would be dropped by later games, but it is not unrecognizable as Zelda. Its more linear structure, and its small quests between dungeons, would influence future entries. Its dungeons are still non-linear and interconnected, with bosses, small keys, locked doors, and key items.

There is no map or compass, but the broad idea remains. In the first dungeon, Parapa Palace, the path to the boss and the path to the candle are locked behind doors, so the player must explore other branches to find keys.

The annoying part is how often keys sit at the end of long hallways, forcing the player to grab one and then walk all the way back. It is the weakest kind of backtracking. Later dungeon loops would solve the same problem more elegantly.

Zelda II takes half steps toward later solutions

Zelda II improves some issues from the first game. Each dungeon has the right number of keys and locks, which makes the spaces feel more self-contained. They can almost be mapped like later Zelda dungeons.

But the solution is still incomplete. Keys can still be used across dungeons, so sequence breaking could potentially create trouble. It is a half step toward the cleaner rule that keys belong to one dungeon only.

The same is true of dungeon items. Most dungeons still allow the player to leave without collecting the item. Only the second dungeon places an obstacle before the boss that forces the player to get the item first.

There is at least a useful completion signal: dungeons turn to ruins on the overworld if the player has both found the item and killed the boss. That helps prevent unnecessary returns to completed areas, but it does not fully solve the skipped-item problem.

The game also still lets players reach dungeons without the outside abilities they need. Miss a somewhat hidden upward thrust technique and one later room becomes a wall. Reach certain bosses without the required spells and the fight is effectively over before it begins.

Combat pressure fights exploration

The dungeons in Zelda II are not especially hard to navigate. They are often small, and the main thing to track is keys and locks. There are few puzzles or obstacles beyond that, and by the time the dungeons grow larger the magic key removes the need to search for keys in the last two palaces.

Players may lose their bearings because many rooms look similar, a problem the first game also had. But a quick mental or paper map is usually enough.

The real challenge is everything around navigation: demanding combat, invisible pits, knockback into lava, pointless dead ends, and a harsh death structure that can send the player all the way back to the starting screen.

That level of punishment makes exploration feel dangerous and frustrating. Later Zelda games would learn an important balance: more linear dungeons can support harder fights, while open and exploratory dungeons need fewer enemies so backtracking does not become exhausting.

Later games refine the foundation

The first two Zelda games set the groundwork for the series, but they also show the mistakes that later games needed to fix.

A Link to the Past and Link's Awakening added stronger puzzles, more distinct dungeons, more memorable individual rooms, and clearer item-based obstacles. Ocarina of Time and Majora's Mask translated those ideas into 3D spaces that could unfold like puzzle boxes.

The handheld entries explored alternate dungeon structures, while later console games often made navigation easier and focused more on individual puzzles, combat, and memorable moments. The DS entries experimented with central dungeons that were revisited across the adventure, and Skyward Sword returned to larger architectural puzzles with places like the Sandship and Sky Keep.

A Link Between Worlds then rethought the formula with a more open dungeon order, setting the stage for an even more radical reinvention later on.

Seen in that long arc, the first two Zelda games are not just historical curiosities. They are prototypes for a dungeon language: brilliant in outline, rough in execution, and essential because every later refinement is easier to understand when the earliest mistakes are visible.