How To Write A Fantasy Novel

The actual problem nobody talks about

Most fantasy novel drafts die because the author spends six months building a magic system and forgets to write a story that uses it. This is not theoretical. I watch this happen constantly in workshop settings and manuscript submissions. You can have the most detailed economic system for a fictional copper currency in a world that uses iron coinage, but if your protagonist doesn't want anything and faces no real obstacles, none of that matters. The core mechanism of writing fantasy fiction is simpler than most guides make it seem. You take a normal human desire, place it inside a setting where the normal rules don't fully apply, and then make things difficult for that character in a way that forces them to change. The fantasy elements are infrastructure. They support the story. They are not the story itself.

How To Write A Fantasy Novel: Starting With The Engine, Not The Painting

Begin with what your protagonist wants and what is stopping them from getting it. The fantasy setting comes after that. When I was drafting my second novel, I had spent approximately three weeks designing a complex system of weather-sorcery where mages could only cast spells during thunderstorms because atmospheric ionization acted as a conductor. I loved the system. It felt clever. My draft hit chapter four and I realized the protagonist had no clear motivation beyond surviving. The magic system was doing all the heavy lifting while the character was essentially standing still. I deleted the first three chapters and rewrote the opening around a character who wanted to keep their younger sibling alive during a famine, and suddenly the weather magic became relevant because only certain regions got rain and control over water distribution meant control over survival. The system finally served the story instead of the other way around. This is a common inversion problem. Writers build the world first and then try to figure out what story fits inside it. It rarely works cleanly. Start with the want and the obstacle. The setting should make those harder, not replace them. When you do build your world, pick your level of detail deliberately. Hard worldbuilding means every element has rules and consequences that are visible to the reader._soft worldbuilding leaves gaps that the reader fills in themselves. Neither is wrong. They are just different tools for different story types. A political intrigue fantasy with factional alliances benefits from hard worldbuilding because readers need to track the rules of succession and trade. A character study about grief set in a magical world works better with soft worldbuilding because the mysteries reinforce the emotional tone. Mixing them carelessly creates reader confusion. If you establish a hard rule in chapter two about how necromancy requires a body part from the deceased, you cannot invoke soft worldbuilding in chapter fifteen to conveniently allow a resurrection with no cost.

One thing beginners consistently miss is the difference between exotic and interesting. Exotic means your setting has weird stuff. Interesting means that weird stuff creates genuine dramatic tension. A market where people trade memories is exotic. A market where a desperate protagonist must sell the memory of their dead parent to buy medicine for their child is interesting. The second line generates from the first, but most writers stop at the first line and call it worldbuilding. Character design follows a similar pattern. The tired expert archetype exists for a reason, and leaning into a character who has seen too much and is barely keeping it together tends to generate stronger prose than the eager orphan who discovers their powers. Not always. But consistently enough that I recommend it as a default starting point. Your protagonist should be competent at something before the story disrupts that competence. A healer who is excellent at their job and then gets cursed with the ability to feel everyone's pain around them creates more narrative engine than a healer who is mediocre and discovers they are special. Competence followed by disruption is more compelling than incompetence followed by discovery.

The drafting process most people get wrong

Write the first draft badly on purpose. This sounds obvious until you sit down and try to make every sentence elegant. I have watched writers produce twelve thousand words in three months because they kept going back to fix paragraph five. The goal of a first draft is not to be good. The goal is to exist. You can fix a broken draft. You cannot fix a blank document. Target length matters less than you think. A fantasy novel sits somewhere between eighty and one hundred twenty thousand words depending on subgenre. Epic fantasy runs longer. Urban fantasy runs shorter. Literary fantasy varies wildly. What matters more is that your story has three structural turns: the inciting incident that disrupts the normal world, the midpoint reversal that changes the direction of the conflict, and the climax where the protagonist either succeeds or fails based on the choices they made throughout. If your midpoint is just another adventure instead of a reversal, your story loses tension. The midpoint should force the protagonist from reactive to active or vice versa. Here is a practical workaround for a specific problem: plot drift during long manuscripts. Around chapter twelve of any fantasy novel above eighty thousand words, I tend to lose track of secondary character arcs and loose plot threads. My solution is a running beat sheet. Not a full outline. Just a single document where I write one sentence per scene describing what happens and which character's want-or-obstacle dynamic it advances. When I read through the sheet and a scene does not connect to a want or obstacle, I cut it or rewrite it. This usually catches drifting plot in about twenty minutes rather than after I have written forty thousand words of dead material.

Dialogue in fantasy has a specific trap. The exposition dialogue problem. Two characters explaining the magic system to each other because the reader needs to know it. This reads like a textbook wearing a costume. The fix is to have characters disagree about what they know. If one character thinks a certain spell is safe and the other knows it is lethal, the information comes out through conflict rather than narration. Natural and efficient. Revision is where the actual writing happens. Your first draft is research. Your second draft is architecture. Your third draft is decoration. Most writers skip from research to decoration and wonder why the house falls down. Do not skip the architecture pass. Read your manuscript once looking only at plot structure. Does every scene raise the stakes? If a scene can be removed without collapsing the story, remove it. I once deleted a entire chapter that was beautifully written, emotionally resonant, and completely unnecessary to the plot. It was the hardest chapter to cut. The book improved immediately after I removed it. Some methods fail under specific conditions. Worldbuilding bibles work well for authors who need reference consistency, but they can become a procrastination tool if you spend more time maintaining the bible than writing the manuscript. I recommend a limit: you may not add a new entry to your bible unless you use that information within the next three chapters. If you write it and never use it, delete it. This keeps your bible lean and your writing productive.

Pacing is another area where beginners make systematic errors. They treat action sequences as the only high-pacing material. Quiet scenes also have pace. A conversation where a character slowly realizes their mentor is lying has pace if the information reveals itself in increments. A battle scene with no emotional stake has no pace regardless of how many swords are swinging. Pace is the rate at which information and consequence arrive. Design it deliberately.

Finishing and beyond

Your ending should resolve the central want-or-obstacle question established in your opening. It does not need to be happy. It needs to be earned. If your protagonist gets what they want without changing, the story is empty. If they change but do not get what they want, the story is tragic. Both are valid. Neither is accidental. Decide which one you are writing before you reach the final act. Querying traditional publishers is a separate discipline from writing. If that is your path, you need a query letter that summarizes your premise in two paragraphs and a synopsis of about five hundred words. These documents require a different skill set than novel writing. Many writers skip this preparation and send raw pitches that editors reject within seconds. Learn the format separately. The practical reality is that most fantasy novels are never published. Not because they are bad, but because the market is saturated and taste is subjective. Writing one anyway is still the only way to learn how to write one. The process itself teaches more than any guide can transfer. You will discover your own patterns, your own weaknesses, and your own voice through the act of finishing something rather than reading about finishing something.

I keep a simple checklist for every draft now: protagonist want, protagonist obstacle, fantasy element integration, three-act structure, rising stakes, and a resolution that matches the genre promise. If one of those boxes is unchecked, the book is not ready. That has saved me from shipping broken manuscripts more times than I can count.