The best fantasy series ask something of you that a novel doesn’t. Ten thousand pages is not a weekend. It’s a year of your reading life, sometimes more, and the genre has a long history of asking for that year and then not delivering an ending.
So most lists of the best fantasy series answer the wrong question. They rank by quality, which you can get from any list of fantasy books, and then leave out the thing you actually need before starting: how long is this, does it finish, does it sag anywhere, and if it isn’t finished, can I stop somewhere without feeling cheated.
This list is organized around those questions. Every entry tells you the length, the completion status, where the weak stretch is if there is one, and whether there’s a safe exit point. Three sections: series that are finished and worth it, one that isn’t finished but has a genuine stopping place, and the ones where you should know exactly what you’re signing up for.
Twenty series. That number isn’t arbitrary; it’s how many I think clear the bar of being worth the commitment. Several famous series aren’t here because popularity isn’t the same as being worth a year, and a few deep cuts are here because this is exactly the page where they pay off. For the wider genre, my list of the best fantasy books covers standalones and the historical canon. If commitment itself is the obstacle, my cozy fantasy books list runs the other way entirely, and the wuxia epics on my Chinese fantasy novels list ask for the same year with very different furniture.
Fiction by René Rehn
Finished, and Worth It
The Lord of the Rings by J. R. R. Tolkien
Three volumes, roughly 480,000 words. Finished in 1955.
The commitment question for Tolkien isn’t length, it’s the first hundred pages.
Fellowship opens slowly, closer in register to English pastoral than to anything the genre became, and readers arriving from modern fantasy series routinely stall in the Shire. The book does not begin behaving like an adventure until Bree, and it isn’t in a hurry to get there.
Push through and the structure justifies itself: the pace accelerates continuously across three volumes, and the final approach through Mordor is as bleak and sustained as anything in the genre. The appendices are optional. The songs can be skipped without guilt, whatever purists say.
One genuine warning: The Silmarillion is not the next book. It’s a mythography, written in a compressed biblical register, and readers who love the trilogy frequently bounce off it hard. Read it if you want the deep history, not because you want more of the same thing. My Tolkien books guide sorts that whole shelf by who it’s for.
The Malazan Book of the Fallen by Steven Erikson
Ten volumes, roughly 3.3 million words. Finished in 2011.
My favorite fantasy series ever written, and the largest commitment on this page by a considerable margin.
Here’s the honest version of what you’re signing up for. Gardens of the Moon drops you into a siege with no explanation, no orientation, and a cast of dozens, and Erikson refuses to hold your hand at any point. He wrote it years before the rest and it reads differently. The standard advice is that the series properly begins with Deadhouse Gates, the second book, and the standard advice is right.
What you get for that difficulty is a world with three hundred thousand years of history behind it, dead civilizations layered under living ones, and myths that turn out to be garbled accounts of things you’ll later see happen. Erikson trained as an archaeologist and an anthropologist and it shows in the texture of everything.
The cast is enormous and somehow it works. Whiskeyjack, Kalam, Quick Ben, Tehol and Bugg, and above all Anomander Rake, one of the great characters in the genre. Most of the leads are soldiers, and the books run on their dry, filthy, exhausted humor.
It’s also more thoughtful than its reputation suggests. Toll the Hounds stops for long philosophical passages on grief and compassion that sit strangely and brilliantly beside the carnage, and the whole sequence resolves toward forgiveness rather than bleakness, which is why the grimdark label never fitted it.
If you want the single best argument for committing: no other fantasy series has this emotional range.
Memory, Sorrow, and Thorn by Tad Williams
Three volumes, roughly 1.1 million words. Finished in 1993.
The trilogy that directly inspired Martin, and a personal favorite without a single reservation.
The commitment problem here is specific and worth naming: Williams spends several hundred pages in the castle before the plot properly starts. Simon is a scullion, he does scullion things, the king dies slowly, and readers who expect a fantasy series to move will find the opening a genuine test.
That patience is the entire design. The back half detonates precisely because the first act established the ordinary world in such detail, and the trek through the winter mountains in the middle volume ranks with anything in the genre for sheer attrition. Nothing in the payoff would land without the slow build.
Pryrates is the thing I’d point at. Most fantasy villains arrive with a rationale; this one is a red-robed priest who wants power, tortures people for information he could obtain more easily, and everyone in the story knows exactly what he is and can do nothing about it.
The invented languages don’t come off, which is the one place the seams show.
Note that the third volume is usually sold split in two, so you’ll see it listed as four books. Williams returned to the world decades later with The Last King of Osten Ard, which is optional and not required for a complete experience.
Discworld by Terry Pratchett
Forty-one novels. Finished in 2015 with Pratchett’s death.
The largest series here and, paradoxically, the smallest commitment, because Discworld is not a series in the way everything else on this page is.
There’s no continuous plot. Books are grouped into loose sequences following different characters, and you can start almost anywhere, read in almost any order, and stop whenever you like without a cliffhanger hanging over you. That’s the answer to the question everyone asks, which is where to begin.
Do not begin at the beginning. The Colour of Magic and The Light Fantastic are parodies of a genre that has since moved on, and they are the weakest books Pratchett wrote. Starting there has put more readers off Discworld than anything else.
Start with a sequence instead. Guards! Guards! opens the Watch books, which are the strongest run and the most conventionally satisfying. Wyrd Sisters opens the witches. Mort opens the Death books. Small Gods and Night Watch are standalone enough to read cold and are the two most often named as the peak.
The series improves for roughly thirty years and then declines sharply at the very end, as Pratchett’s Alzheimer’s progressed. The final books are difficult to read for reasons that have nothing to do with quality. The sub-series map with entry points is on my Terry Pratchett guide.
The First Law by Joe Abercrombie
Three volumes plus three standalones plus a further trilogy. The core trilogy finished in 2008.
The most efficient commitment on this page, and the one with the clearest exit points.
The original trilogy is self-contained and resolves. You can stop there having read a complete story, and many people do. What follows is three standalone novels set in the same world, of which Best Served Cold is the strongest, and then the Age of Madness trilogy set decades later, which is genuinely good and entirely optional.
That modular structure is unusual in fantasy series and it’s a real advantage: the commitment is three books, and everything after is opt-in. The full modular structure is laid out on my Joe Abercrombie guide.
Fair warning on tone. The trilogy is built to subvert the expectations it sets up, the quest achieves nothing anyone intended, and readers who want the traditional payoff will find the ending an anticlimax by design. Glokta’s chapters are the reason most people finish it.
Earthsea by Ursula K. Le Guin
Six books, all short. Finished in 2001.
The shortest complete series here, and the one that changes its mind about itself halfway through.
The first three books are a compact trilogy written in the late sixties and early seventies: a wizard’s education, a girl in a labyrinth, a voyage to the edge of the world. Each is under 250 pages and they can be read in a week.
Then Le Guin returned eighteen years later and wrote Tehanu, which looks back at everything the trilogy assumed and asks who was left out of it. Wizards are men. Women get village witchcraft, considered lesser. The earlier books never questioned that and this one does, at length.
The commitment question is therefore unusual. Read as three books, it’s a perfect small trilogy. Read as six, it’s a writer arguing with her younger self across two decades, which is more interesting and considerably less comfortable. Tehanu is slower and more domestic than what precedes it, and some readers feel it undermines what they loved.
Both are legitimate ways to read it. Just know the shift is coming.
The Book of the New Sun by Gene Wolfe
Four volumes, individually short by epic standards. Finished in 1983, with a fifth coda volume.
The hardest thing on this page and the one its admirers rate highest in the genre.
Severian is a torturer’s apprentice exiled from his guild, traveling across a far-future Earth where the sun is dying and the technology left behind is indistinguishable from magic. He tells you at the outset that he has a perfect memory. He is lying, or mistaken, or deliberately concealing things, and working out which is the actual activity of reading the book.
The commitment here isn’t length, it’s density. Wolfe buries plot in offhand clauses, uses archaic vocabulary for real things rather than invented ones, and expects you to notice when the narrator’s account contradicts itself. Most readers who love it have read it more than once, and many say the second reading is where it becomes a different book.
The fifth volume, Urth of the New Sun, is a coda written years later and is genuinely optional.
If you want the genre’s most demanding series and you have the patience for it, this is the summit. If you want to be carried along, start elsewhere.
Realm of the Elderlings by Robin Hobb
Sixteen books across five sub-series. Finished in 2017.
The longest emotional commitment in fantasy, and the one most likely to leave a mark.
It opens with the Farseer trilogy: FitzChivalry, a royal bastard nobody wanted, is taken in by the court and shaped into its assassin. Hobb writes him in a tight first person across decades, and the defining quality of the whole sequence is that suffering accumulates rather than resetting between books.
The structure matters for planning. Farseer is a complete trilogy and a legitimate stopping point. Liveship Traders that follows has different characters entirely and is the best-written of the sub-series. Tawny Man returns to Fitz. Then Rain Wilds, which is the weak stretch and the one most readers say can be skipped. Then Fitz and the Fool, which closes everything.
The ending is famously devastating and divides readers permanently between those who consider it perfect and those who feel betrayed by it.
Be warned that Hobb is relentless with her protagonist. If you want a series where things go well for people you like, this is not it. The sub-series order and the skip question are handled on my Robin Hobb guide.
The Broken Earth by N. K. Jemisin
Three volumes. Finished in 2017.
The tightest complete trilogy here, and the only series to win the Hugo for all three volumes.
The commitment is genuinely small, three books of ordinary length, and the payoff is unusually high. A supercontinent wracked by geological catastrophes so regular that civilization has rebuilt itself around the certainty of the next one, and a caste of people who can control seismic energy and are enslaved for it.
One structural device carries the whole thing and I won’t describe it. What’s worth knowing is that one of the three narrative threads is written in second person, which some readers find alienating for the first fifty pages and which turns out to be load-bearing.
Frequently shelved as science fiction, and reasonably so, since the orogeny has a physical basis and the explanation when it arrives is technological.
If you want proof that a fantasy series doesn’t need ten volumes, this is the argument.
The Black Company by Glen Cook
Ten books in the main sequence. Finished in 2000.
The series working fantasy writers cite most often as an influence, and the one with the most distinctive prose problem.
The conceit is that you’re reading a mercenary company’s official annals, written by its physician, Croaker. He records events and omits interpretation, occasionally admitting he’s shaped things for the men’s morale. The result is deliberately flat: no scenery, minimal interiority, atrocities logged as line items.
That flatness is the achievement and it’s also the reason people bounce. Cook is not going to tell you how to feel about anything, and readers who want emotional guidance from a narrator will find the first book cold.
The commitment breaks into segments. The first three, collected as Chronicles of the Black Company, form a complete arc and are the strongest. The Books of the South follow. The Glittering Stone sequence closes it. Quality varies more than most series here, and the consensus is that the opening trilogy is the essential part.
Grimdark as a category descends more or less directly from these books.
The Gormenghast Novels by Mervyn Peake
Three volumes, though effectively two. Ended by Peake’s illness in 1959.
The most beautiful prose on this page and the most complicated completion status.
Titus Groan and Gormenghast are a matched pair and a complete experience: a castle too vast for its own inhabitants to map, run on ceremonies nobody has understood for centuries but everyone must perform, and a kitchen boy named Steerpike who sees that a structure this rigid is a ladder for anyone willing to kill.
Then there’s Titus Alone, written while Peake was declining from a degenerative illness, and it is a genuinely different and much weaker book. The setting changes, the register changes, and the editorial history is contested.
So the honest commitment answer is two books. Read Titus Groan and Gormenghast as a finished work, because they are one. Treat the third as an appendix for people who want to see where he was heading. The man and the rest of his work are covered on my Mervyn Peake guide.
Peake trained as an illustrator, and it shows in prose that draws every face and room rather than describing them. It’s slow by design, and readers quit early or convert completely.
The Green Bone Saga by Fonda Lee
Three volumes. Finished in 2021.
The best-structured recent trilogy, and the one that handles time best.
On the island of Kekon sits on the world’s only jade, the mineral that makes trained fighters superhuman, and the clans guarding it function as family, business, and government at once. The model is explicitly the gangster epic, and the Kaul family’s war with its rival plays out over decades rather than months.
The generational structure is why it’s here. Each volume jumps years forward, characters age, children become players, and the consequences of book one’s decisions land in book three on people who weren’t born yet. Very few fantasy series attempt that and fewer land it.
It also ends properly, which after several entries on this page is worth stating plainly. Lee planned three books, wrote three books, and the finale resolves the family’s arc without leaving threads for a sequel.
Modern setting, no medieval furniture, and a good entry point for someone who thinks they don’t like fantasy series.
The Broken Empire by Mark Lawrence
Three volumes. Finished in 2013.
The shortest serious commitment here and the most divisive.
Jorg Ancrath is thirteen when the trilogy opens and already leads a crew of killers, and Lawrence writes him in a compressed, aphoristic first person that is genuinely brilliant and genuinely uncomfortable. You spend three books being entertained by someone doing indefensible things, which is the point and which plenty of readers refuse.
The commitment advantage is real: three ordinary-length books, fully resolved, with a reveal about the nature of the world that recontextualizes the whole trilogy.
Lawrence has written a further trilogy in the same world, The Red Queen’s War. It’s better-behaved and less interesting.
Know what the opening involves before you start. If a protagonist you actively despise sounds unbearable rather than compelling, this isn’t your series.
The Bas-Lag Novels by China Miéville
Three books. Effectively finished, though Miéville has never ruled out returning.
A qualified entry, because Miéville isn’t really a fantasy writer and these aren’t really a series.
The three books share a world and nothing else. No continuing protagonist, no ongoing plot, no reading order that matters much, though Perdido Street Station first makes sense. Each is a complete novel, which means the commitment is unusually low: read one, and if the sensibility works for you, read the others.
What you get is New Crobuzon, an industrial city running on steam, thaumaturgy, and exploitation, with the best monsters in modern fantasy in the slake-moths and one of the genre’s great images in the perpetual train of Iron Council.
The honest framing is that Miéville belongs to weird fiction rather than to fantasy proper, and his project is closer to Borges and the New Weird than to anything else on this page. The fantasy furniture is there and it’s being used for something else entirely. I’ve covered that tradition separately, because it deserves treating on its own terms rather than as a fantasy subgenre.
Included here because readers looking for fantasy series find their way to Bas-Lag constantly, and they should know what they’re getting.
Under Heaven and the Kay Sequence by Guy Gavriel Kay
Standalones and loose pairs rather than a true series. All complete.
The entry for readers who want everything on this page except the commitment.
Kay writes historical fantasy: a real period, lightly fictionalised, with a thread of the supernatural running through it. Under Heaven and River of Stars are Tang and Song dynasty China. The Lions of Al-Rassan is Moorish Spain. The Sarantine Mosaic is Byzantium and is the closest thing he has to a two-book series.
Nothing here requires reading anything else. That’s the appeal, and it’s why he’s on a list about commitment: Kay is where you go when you’ve finished a ten-volume epic and cannot face another one.
The Lions of Al-Rassan is the usual entry point and the one most readers name as his best. The prose is more consciously literary than the genre standard, and the emotional register is elegiac almost throughout.
Worth knowing that his early Fionavar Tapestry trilogy is conventional portal fantasy and unrepresentative of everything after.
Unfinished, With a Real Stopping Point
The Stormlight Archive by Brandon Sanderson
Five of a planned ten volumes, each enormous. Arc one complete as of 2024.
The largest work in progress in the genre, and the only unfinished series on this page with a genuine exit.
Sanderson planned ten books as two five-book arcs, and the fifth volume closes the first arc deliberately. Major plotlines resolve, the central conflict reaches a conclusion, and a reader who stops there has a complete experience rather than a cliffhanger. That’s a real structural promise and he kept it.
The commitment is still enormous. The books average around four hundred thousand words, with the later ones running well past it, which means arc one alone is longer than most completed series on this page.
What you get for it is the most rigorous world-building in modern fantasy, and something the genre rarely manages well: Kaladin’s depression and Shallan’s dissociation are written with real seriousness and are structural to the plot rather than decorative.
Sanderson also finishes what he starts, on schedule, which after the next section will look like a considerable virtue. The reading order for everything around it is on my Brandon Sanderson guide.
Unfinished, Know What You’re Signing Up For
A Song of Ice and Fire by George R. R. Martin
Five of a planned seven volumes. The most recent was published in 2011.
I read to the fourth book and loved it, and I’d tell anyone starting now to go in with clear eyes.
What’s there is genuinely excellent. The pitch takes one sentence: a dynastic war over a single throne while a much older threat gathers beyond the northern wall. The low magic is what sets it apart, the supernatural rationed so strictly that every appearance feels like an event, and the foreground given over to politics.
The point-of-view structure is the achievement. The viewpoint structure hands chapters to heroes and monsters alike until the labels stop holding, which is why the books outlive their own television adaptation in readers’ memories.
The problem is obvious and worth stating plainly. It has been unfinished for well over a decade, the announced sixth volume has no date, and it is entirely reasonable to assume it will stay that way. There is no safe stopping point; book five ends mid-arc for multiple characters.
Read it as a magnificent incomplete thing, or don’t read it. Both are defensible. The rest of Martin, including the career almost nobody reads, is on my George R. R. Martin guide.
The Wheel of Time by Robert Jordan
Fourteen volumes plus a prequel. Completed in 2013 by Brandon Sanderson after Jordan’s death.
Technically finished, listed here because the commitment carries a specific and well-documented risk.
I’ve read the first book. Eye of the World is a competent, openly Tolkien-shaped opening: a farm boy, a dark threat, a wise stranger, a road. It does what it does well and it gave me no particular urge to commit to thirteen more.
What the consensus says about the rest is consistent enough to be useful. The series is genuinely great through roughly book six, then slows severely across the middle volumes, with books eight through ten frequently described as the point where plot progress nearly stops. Then Sanderson took over from Jordan’s notes and the final three accelerate hard to a well-regarded ending.
So the honest commitment picture is: well past four million words, a genuinely excellent first third, a middle section that a substantial number of readers abandon, and a strong finish that requires getting through it. Reading orders exist online specifically to help people survive the middle.
It’s foundational to modern epic fantasy and it is a very large ask.
The Kingkiller Chronicle by Patrick Rothfuss
Two of a planned three volumes. The second was published in 2011.
The most beautifully written unfinished series in fantasy, and the one I’d think hardest about before starting.
Kvothe narrates his own life to a chronicler across two enormous books: an orphaned childhood, a magical university, a legend accumulating around him. The prose is genuinely superb and the frame narrative is doing something clever, since the man telling the story is diminished in a way the story he tells doesn’t explain.
The second volume also contains a long digression that many readers consider the point where the structure lost its discipline, and it’s the most common reason people cool on the series.
Fifteen years without a third book. No announced date. Rothfuss has been publicly non-committal in a way that has exhausted much of the fanbase, and the two published volumes resolve almost nothing.
If you can enjoy two thousand pages of setup for a payoff that may never arrive, it’s remarkable writing. If that sounds like a bad trade, it is one. The whole situation gets an honest accounting on my Patrick Rothfuss guide.
The Gentleman Bastards by Scott Lynch
Three of a planned seven volumes. The most recent was published in 2013.
The most purely entertaining series here and the one with the least likely completion.
Locke Lamora is a con artist in Camorr, a Venice-like city of canals and elaborate crime, and the first book alternates between an ingenious heist and the childhood that produced him. The dialogue is superb, the plotting is tight, and the friendship at the center is written with genuine warmth, which is why the violence lands.
The Lies of Locke Lamora is a complete novel with a satisfying ending, which makes it a legitimate one-book read. That’s the practical recommendation: treat book one as a standalone.
The two sequels are good and progressively less self-contained, and the third ends on threads that have now been open for over a decade. Lynch has spoken openly about the health difficulties behind the delay, which is a better reason than most, and it doesn’t change the reader’s situation.
If you’re starting from nothing: The Broken Earth if you want three books and an ending, First Law if you want the modular version, Malazan if you want the biggest thing the genre has to offer, and Discworld if you’d rather not commit to anything at all.
