What is the best build for Karthus?

We do not publish a best build for Karthus, and the reason is worth thirty seconds. The most recorded Karthus jungle build on patch 16.14 is Blackfire Torch, Zhonya's Hourglass and Shadowflame, in 44 games. A finished item set is read at the end of the game, so a build only appears once the game ran long enough to buy it, and a game that runs long is usually a game you were winning. Any share of wins taken from those games leans high for that reason on its own, which is why there is not one in this answer. The page lists every recorded build with the games behind it.

What are the best runes for Karthus?

The most recorded Karthus jungle rune page on patch 16.14 takes Dark Harvest out of Domination, with Cheap Shot, Sixth Sense and Treasure Hunter beside it and Last Stand and Triumph out of Precision. Players who ran it won 50.7% of 286 games, range 44.9% to 56.4%. A rune page is locked in at champion select, so that share is not inflated by long games the way a build is, but it counts who picked the page rather than what the page did.

What skill order should I use on Karthus?

Recorded Karthus jungle players max Q, then E, then W. 146 recorded games on patch 16.14.

Karthus build, patch 16.14

1,165 games of Karthus recorded on patch 16.14, across jungle, bot, top, mid and support.

Win rate

47.7% of jungle games this patch

Pick rate

0.7% of jungle games

Games

555 jungle, patch 16.14

Rank in role

65 of 77 jungle champions

Win impact against the average jungle pick on patch 16.14: -0.5 wins per 100 games (range -2.4 to 1.4) 555 games Pooled

What most Karthus jungle players build, patch 16.14

Build Blackfire Torch, Zhonya's Hourglass and Shadowflame, and run Dark Harvest

The rune page they run

Domination

Dark Harvest

Precision

Shards

286 games ran this page exactly as it is drawn. We never stitch the most picked keystone onto the most picked second row and call it a page: nobody ran that one.

The skill path Each filled cell is the skill most players took at that level, from 146 games.
skill	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18
Q	Q		Q		Q		Q		Q
W				W										W	W		W	W
E		E						E		E		E	E
R						R					R					R

Levels 10 to 18

10 E, 11 R, 12 E, 13 E, 14 W, 15 W, 16 R, 17 W, 18 W

Max Q, then E, then W. The level by level count.

44 games finished this exact set. It is the set recorded most often, which is a fact about players rather than about the items.

Share of those games that were wins, in percent 57% (range 42-70) · 44 games · Measured

Games that finish a build are games that lasted, and lasting usually means winning, so this figure leans high before the items do anything. It is how those games went, not what the build does to your chances. How far that can bend a number.

Karthus jungle is the role with the most recorded games on patch 16.14: 555 games, of which 265 were wins. That is the role this page leads with. Measured on this page: jungle, bot, top and mid, the roles that cleared 30 recorded games.

wins 0.5 per 100 vs the jungle average · 555 games on 16.14 · first-time players measured (500 games) · no clear move vs 16.13 (-10.0, range -24.8 to +4.8)

The build above is the one recorded most often, nothing more. Every recorded build, with its own games and its own share, is in the build list, in the order players built them.

Why this answer

A champion and role pair needs 30 recorded games before we put a score or a win impact on it: under that, a handful of lucky games moves a win rate by ten points. The same floor applies to a single build and to a single rune page, and anything under it prints its games and no number.

The share on that figure says how the games went, not what the build does to you. Nothing on this page measures that; the method note at the end says why.

What do most Karthus players build?

Item sets recorded on patch 16.14, most built first, never sorted by how the games went. The share is how many of a build's own games were wins; a build under 30 games shows its games and no share.

Karthus, jungle

8 item sets were recorded in this role and 4 of them cleared 30 games.

Blackfire Torch, Zhonya's Hourglass and Shadowflame
56.8% 44 games
Blackfire Torch, Rabadon's Deathcap and Shadowflame
47.6% 42 games
Blackfire Torch, Rabadon's Deathcap, Zhonya's Hourglass and Shadowflame
54.3% 35 games
Blackfire Torch, Shadowflame and Liandry's Torment
51.6% 31 games

A finished set is read at the end of the game, so every share here leans high: games that last long enough to finish a build are usually games that were already going well. How far that can bend a number.

Karthus, bot

8 item sets were recorded in this role and 2 of them cleared 30 games.

Blackfire Torch, Rabadon's Deathcap and Shadowflame
43.5% 46 games
Blackfire Torch, Rabadon's Deathcap, Void Staff and Shadowflame
55.9% 34 games

Karthus, top

0 item sets were recorded in this role and 0 of them cleared 30 games.

No top build recorded yet

We have 63 games of Karthus top on patch 16.14, but no item set repeated often enough to record as a build.

Read the rune pages instead

Karthus, mid

0 item sets were recorded in this role and 0 of them cleared 30 games.

No mid build recorded yet

We have 52 games of Karthus mid on patch 16.14, but no item set repeated often enough to record as a build.

Read the rune pages instead

What runes do most Karthus players run?

Whole pages that players actually ran, most recorded first. A page is locked in at champion select, so long games do not inflate its share the way they inflate a build's; it still counts who picked the page rather than what the page did.

Karthus, jungle

7 complete pages were recorded in this role and 3 of them cleared 30 games.

Dark Harvest Domination with Precision

286 games

Domination

Dark Harvest

Precision

Shards

286 games ran this page exactly as it is drawn. We never stitch the most picked keystone onto the most picked second row and call it a page: nobody ran that one.

Other pages players ran

Deathfire Touch Sorcery with Precision

161 games

Dark Harvest Domination with Sorcery

66 games

Keystones picked, one slot on its own

Dark Harvest: 48.3% 354 games
Deathfire Touch: 47.9% 190 games

3 keystones were recorded in this role: one slot on its own, counted across every page that used it, so the share carries the rest of each page with it.

Karthus, bot

4 complete pages were recorded in this role and 2 of them cleared 30 games.

Deathfire Touch Sorcery with Precision

358 games

Sorcery

Deathfire Touch

Precision

Shards

358 games ran this page exactly as it is drawn. We never stitch the most picked keystone onto the most picked second row and call it a page: nobody ran that one.

Other pages players ran

Dark Harvest Domination with Precision

110 games

Keystones picked, one slot on its own

Deathfire Touch: 53.1% 360 games
Dark Harvest: 48.8% 127 games

3 keystones were recorded in this role: one slot on its own, counted across every page that used it, so the share carries the rest of each page with it.

Karthus, top

4 complete pages were recorded in this role and 1 of them cleared 30 games.

Deathfire Touch Sorcery with Precision

43 games

Sorcery

Deathfire Touch

Precision

Shards

43 games ran this page exactly as it is drawn. We never stitch the most picked keystone onto the most picked second row and call it a page: nobody ran that one.

Keystones picked, one slot on its own

Deathfire Touch: 63.8% 47 games

3 keystones were recorded in this role: one slot on its own, counted across every page that used it, so the share carries the rest of each page with it.

Karthus, mid

3 complete pages were recorded in this role and 1 of them cleared 30 games.

Deathfire Touch Sorcery with Precision

44 games

Sorcery

Deathfire Touch

Precision

Shards

44 games ran this page exactly as it is drawn. We never stitch the most picked keystone onto the most picked second row and call it a page: nobody ran that one.

Keystones picked, one slot on its own

Deathfire Touch: 48.9% 45 games

2 keystones were recorded in this role: one slot on its own, counted across every page that used it, so the share carries the rest of each page with it.

What skill order should I use?

Counted from recorded level-ups on patch 16.14, one player per game.

Max Q, then E, then W

146 recorded games on patch 16.14.

Karthus jungle skill path Each filled cell is the skill most players took at that level, from 146 games.
skill	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18
Q	Q		Q		Q		Q		Q
W				W										W	W		W	W
E		E						E		E		E	E
R						R					R					R

Levels 10 to 18

10 E, 11 R, 12 E, 13 E, 14 W, 15 W, 16 R, 17 W, 18 W

Max Q, then E, then W

134 recorded games on patch 16.14.

Karthus bot skill path Each filled cell is the skill most players took at that level, from 134 games.
skill	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18
Q	Q		Q		Q		Q		Q
W				W										W	W		W	W
E		E						E		E		E	E
R						R					R					R

Levels 10 to 18

10 E, 11 R, 12 E, 13 E, 14 W, 15 W, 16 R, 17 W, 18 W

Karthus, top

Not enough top level-ups yet

not enough games yet. 16 games so far, this needs 30.

Karthus top has 63 games recorded, but only 16 games of them carry level-ups. A levelling order off that many games is luck.

How we count level-ups

Karthus, mid

Not enough mid level-ups yet

not enough games yet. 18 games so far, this needs 30.

Karthus mid has 52 games recorded, but only 18 games of them carry level-ups. A levelling order off that many games is luck.

How we count level-ups

When do players buy each piece?

The pieces of the most recorded Karthus jungle build, in the order players buy them. The time is the middle of the spread across every recorded game on patch 16.14, measured at the buy rather than at the end of the game: half the recorded buys of the item happen before it, half after.

Buy times, earliest first

Blackfire Torch: 11:08 1,142 games
Shadowflame: 20:43 1,918 games
Zhonya's Hourglass: 25:56 1,680 games

The bar is scaled to the latest piece of the set. Each item links to its own page, where the whole spread sits beside the games behind it.

Who counters Karthus?

The opponents Karthus jungle loses to most and the teammates Karthus gains the most beside on patch 16.14. The chip is the edge in wins per 100 games, and the games behind it sit under the name.

Hardest opponents

Not enough games yet to rank Karthus's opponents. Karthusjungle has 574 games recorded on patch 16.14, and ranking opponents needs 1,000 before the baseline is steady; every opponent so far is counted, not ranked.

Best teammates

Not enough games yet to rank the champions Karthus wins beside.

These numbers come from 108,616 ranked games we recorded on EUW, Korea and NA, found by following Diamond players and the people they play with. That is not every rank and it is not every region. Everything on this page is patch 16.14. Counted up to 24 July 2026.

How does Karthus rank?

Every role we recorded Karthus in on patch 16.14, scored or not. Win impact is measured against the average pick in that role on the same patch, and a role needs 30 recorded games before it is scored.

Karthus by role

jungle C -0.5 wins per 100 games (range -2.4 to 1.4) 555 games Pooled: 43 555 games
bot B +0.4 wins per 100 games (range -1.5 to 2.4) 487 games Pooled: 50 487 games
top C +0.3 wins per 100 games (range -1.8 to 2.5) 63 games Pooled: 47 63 games
mid C 0.0 wins per 100 games (range -2.2 to 2.1) 52 games Pooled: 45 52 games
support not enough games yet. 8 games so far, this needs 30.: not scored 8 games

Why no share of wins on this page says a build is good for Karthus

A finished item set is read at the end of the game. A player who bought the first two items and lost at twenty minutes never finished the set, so they are not in its row. Everybody in the row lasted long enough to buy it, and lasting is most of what winning looks like. The share those games won is pushed up by that on its own, before the items do anything at all, and the push is largest on the most expensive builds and backwards on the ones you buy when a game is going badly.

We know the size of it. In simulation, with every item's real effect set to exactly zero, an item read at the end of the game shows about 92% as a sixth completed item. A defensive item with a real 5 point benefit shows about 20 points the wrong way, because it gets bought when a player is behind. The measurement is here.

So this page counts and labels rather than ranks. The estimate that answers the real question is taken at the moment of the buy, from the state the purchase happened in, and it runs in the desktop app because it needs the game in front of it.