The 50 biggest Korean titles on Netflix’s global weekly chart, out of 243 that have appeared since 2021-07-04. Ranked once by how long they lasted and once by how hard they landed — the two lists are almost different lists.
Of the eight on each side, 4 appear on both — Squid Game and Queen of Tears and The Glory and All of Us Are Dead. A title that runs for months and a title that consumes a fortnight are both “hits”, and the single number usually quoted for either one cannot tell them apart.
| Group | Titles | Median hours | Median weeks | Avg hours |
|---|---|---|---|---|
| Peaked at #1 | 65 | 103m | 5 | 234m |
| Peaked below #1 | 178 | 25m | 2 | 51m |
All 243 titles, not the fifty largest. The median title that reached number one drew 4.2× the hours of one that did not, and it also ran 5 weeks against 2. Most of that gap is length, not landing: hold the run length fixed and what remains is 1.26× among titles that lasted six to ten weeks.
| Titles that lasted | Reached #1 | Median hours | Did not | Median hours |
|---|---|---|---|---|
| 1–2 weeks | 1 | 126m | 90 | 6m |
| 3–5 weeks | 42 | 79m | 57 | 46m |
| 6–10 weeks | 15 | 163m | 25 | 129m |
| 11 weeks or more | 7 | 654m | 6 | 331m |
Neither number is the cause of the other. Weeks on chart and peak rank track total hours about equally well (0.73 and 0.754 against log hours), and they track each other at 0.473, so a table that shows one is partly showing the other. Reaching number two goes with a higher median than reaching number one (123m against 103m) — the order does not run down the ranks. The full story →
53 of 243 Korean titles that reached a global Non-English Top 10 — 21.8% — were on it for exactly one week. Half the panel was gone within two.
| Weeks on chart | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 |
|---|---|---|---|---|---|---|---|---|---|---|
| Titles | 53 | 38 | 35 | 38 | 26 | 7 | 12 | 12 | 5 | 4 |
The tail runs past this table — the longest run in the panel is 42 weeks — but it is thin, and that thinness is the point. A chart run is not a duration most titles have.
What is not on this list. Netflix files its global charts by a title’s primary language, and a Korean work belongs on the Non-English side. 53 titles matched a Korean work by name but sit on the English charts, so they are not that work and are out. 15 more came out after we read them one at a time — The East Palace, Teach You a Lesson, Breathless among them. We hand-checked the largest 30 series and 20 films by hours and stopped there; below that line the language rule is doing the work on its own.
| Title | Type | Weeks | Peak | Hours | Per week |
|---|---|---|---|---|---|
| Squid Game | series | 42 | #1 | 5,048m | 120m |
| Extraordinary Attorney Woo | series | 21 | #1 | 662m | 32m |
| All of Us Are Dead | series | 11 | #1 | 660m | 60m |
| Queen of Tears | series | 15 | #1 | 654m | 44m |
| The Glory | series | 13 | #1 | 561m | 43m |
| Bon Appétit, Your Majesty | series | 10 | #1 | 512m | 51m |
| When Life Gives You Tangerines | series | 9 | #1 | 482m | 54m |
| King the Land | series | 11 | #1 | 447m | 41m |
| Alchemy of Souls | series | 27 | #3 | 419m | 16m |
| Agent Kim Reactivated | series | 8 | #1 | 365m | 46m |
| Bloodhounds | series | 10 | #1 | 353m | 35m |
| Love Next Door | series | 11 | #2 | 340m | 31m |
| My Demon | series | 11 | #2 | 331m | 30m |
| Dynamite Kiss | series | 9 | #1 | 315m | 35m |
| Hometown Cha-Cha-Cha | series | 16 | #2 | 301m | 19m |
| When the Phone Rings | series | 8 | #2 | 282m | 35m |
| Business Proposal | series | 13 | #1 | 279m | 21m |
| Can This Love Be Translated? | series | 5 | #1 | 271m | 54m |
| My Royal Nemesis | series | 8 | #1 | 248m | 31m |
| Genie, Make a Wish | series | 5 | #1 | 242m | 48m |
| True Beauty | series | 7 | #5 | 242m | 35m |
| Gyeongseong Creature | series | 8 | #2 | 204m | 26m |
| Beyond the Bar | series | 7 | #2 | 202m | 29m |
| My Name | series | 5 | #2 | 194m | 39m |
| Weak Hero | series | 7 | #1 | 185m | 26m |
| The Trauma Code: Heroes on Call | series | 5 | #1 | 185m | 37m |
| Crash Course in Romance | series | 11 | #2 | 183m | 17m |
| Strong Girl Nam-soon | series | 9 | #3 | 174m | 19m |
| The WONDERfools | series | 5 | #2 | 171m | 34m |
| Little Women | series | 11 | #3 | 164m | 15m |
| Hellbound | series | 6 | #1 | 163m | 27m |
| Miss Night and Day | series | 10 | #2 | 160m | 16m |
| Doctor Slump | series | 9 | #1 | 154m | 17m |
| Doctor Cha | series | 7 | #2 | 151m | 22m |
| Welcome to Samdal-ri | series | 8 | #5 | 150m | 19m |
| The Great Flood | film | 7 | #1 | 149m | 21m |
| Juvenile Justice | series | 8 | #1 | 147m | 18m |
| Destined with You | series | 8 | #3 | 146m | 18m |
| Sold Out on You | series | 7 | #1 | 142m | 20m |
| Our Unwritten Seoul | series | 6 | #3 | 141m | 23m |
| Resident Playbook | series | 7 | #3 | 140m | 20m |
| The King's Affection | series | 10 | #2 | 139m | 14m |
| Money Heist: Korea - Joint Economic Area | series | 5 | #1 | 134m | 27m |
| As You Stood By | series | 5 | #1 | 133m | 27m |
| Narco-Saints | series | 5 | #1 | 130m | 26m |
| Our Beloved Summer | series | 8 | #4 | 129m | 16m |
| Tastefully Yours | series | 6 | #2 | 128m | 21m |
| The Price of Confession | series | 4 | #1 | 128m | 32m |
| Our Sticky Love | series | 2 | #1 | 126m | 63m |
| The Guest | series | 4 | #6 | 125m | 31m |
Of the ten most watched Korean titles on Netflix's global chart, nine reached number one. Alchemy of Souls charted longer than eight of them — 27 weeks — and peaked at three. Hometown Cha-Cha-Cha lasted 16 weeks and peaked at two.
1 September 2026
Eight months without the top place leaving Korean television, and it changed hands thirteen times inside that run. Malaysia and Singapore then did seventeen weeks each, starting and ending on exactly the same dates.
30 August 2026
Measured on all 246 Korean titles instead of the fifty largest, the number-one group runs five weeks to everyone else's two. Hold length fixed and the gap that is left is 1.3× among titles that lasted six to ten weeks.
8 August 2026 · corrected
215 of 421 titles never reached a global Top 10, so Netflix's language label — the only thing that separates a Korean work from a foreign one with the same name — cannot be applied to them. They are also the small ones.
7 August 2026
Our Korean Netflix catalogue was 294 titles and 27.7 billion hours. It is 236 and 23.7 billion. Here is the mistake, how it spread, why counting alone would never have caught it, and what we changed so it cannot happen the same way twice.
7 August 2026 · corrected
Five years of Netflix's global Top 10 hold 23.7 billion hours of Korean viewing across 236 titles. One of them is a fifth of it. Every title listed.
6 August 2026 · corrected
Fourteen seconds, read aloud: Vietnam's number one series on Netflix was Korean for 34 weeks straight — but not one show. Eleven different Korean series held that place, and it changed hands thirteen times without once leaving Korean television. This counts the top place week by week, not one title's own run, and a rank is an ordering: Netflix publishes no viewers or hours by country. Free to repost with the address on it.
All our short films — each one built from a table on this site
We report the median rather than the mean. A handful of very large values would pull a mean away from where most of the sample actually sits, and in reads-per-title data those large values are the norm, not an error. Correlations are Pearson coefficients. Where hours are involved they are taken on the logarithm of hours, because viewing hours span several orders of magnitude and a handful of enormous titles would otherwise decide the coefficient on their own. Titles are ranked by total viewing hours across the whole window, and staying power is the number of distinct weeks a title appeared on any country chart, so a title that returns after a gap is credited for both spells.
A median tells you where the middle sits and nothing about the shape around it, so it should be read next to the range or the full distribution, not alone. A correlation is not a cause, and it only detects the straight-line part of a relationship — two things can move together tightly in a curve and still score near zero. Taking logs changes what is being correlated: it is the ranking-like structure of hours, not hours themselves. A coefficient from a chosen sample also inherits that choice; ours are titles that reached a chart, which is not a sample of titles. A week on a chart is not a week of watching — a title can be watched steadily and never chart, and the chart cuts off at ten. The window opens in 2021, so titles already running before then start mid-life. And every title here reached a chart at least once, so nothing on this page speaks for titles that never did.