Stephen Mitchell's Tao Te Ching came out in 1988 and has probably introduced more English speakers to the book than every scholarly edition combined. It is calm, spare, quotable, and it moves. If you have seen a line from the Tao Te Ching on a wall, it was likely his.
It is also not a translation, and he has never claimed it was. The book is labelled a "version". Mitchell does not read Classical Chinese; he worked from existing translations and scholarly commentaries, filtered through fourteen years of Zen training, and produced English poetry from them. Knowing that changes what the book is for rather than making it worthless.
The objection, stated fairly
Sinologists have been consistently unhappy, and the complaints are specific rather than snobbish. Passages get smoothed into a serenity the Chinese does not have. Politically pointed chapters — and a good deal of the Tao Te Ching is advice to rulers about governing — come out as personal spiritual counsel. Difficult lines that scholars argue over arrive fully resolved, with the argument invisible. And the gendered pronouns were altered to alternate he and she, a defensible modern choice presented as though it were in the source.
The deeper problem is that a reader cannot tell which lines are Laozi and which are Mitchell, because nothing marks the seam. That is fine for a poem and a genuine problem for a source text.
What it does better than the scholarly editions
It is readable at 7am, which sounds like faint praise and is not. The Tao Te Ching is a book people give up on, and the most common reason is that the accurate translations are dense with commentary, brackets and alternative readings that break the spell the book depends on. Mitchell's has an unbroken voice, and for a first pass — or for reading a chapter a day rather than studying it — that voice is the thing that keeps people coming back.
The other honest defence: the Tao Te Ching has been reinterpreted by every generation that received it for 2,400 years, including in Chinese. A version is part of that tradition, provided it is labelled, and his is.
What to read it alongside
The practical answer is two books, which is cheaper than it sounds.
Addiss and Lombardo is the usual first recommendation for accuracy without deadness: spare, close to the Chinese, with the characters included. D. C. Lau is the standard scholarly English text. Red Pine (Bill Porter) gives you the Chinese commentators across two millennia alongside the text, which is the closest thing to seeing how the book has actually been read. Ursula K. Le Guin published what she also honestly called a rendition rather than a translation, and hers is the direct comparison for Mitchell, done with more disclosure of method.
Our full translation comparison weighs all of them, and which edition to buy covers the physical books.
The verdict
Buy it if you want to read the Tao Te Ching. Do not cite it if you want to argue about what the Tao Te Ching says. Those two sentences cover almost every reader, and the second one is why the scholars are annoyed rather than merely sniffy.
If what you actually want is the daily habit rather than the scholarship, 365 Days of Tao (ours, flagged) gives one verse a day with a plain-English reflection and no pretence that the reflection is Laozi. And how to read the Tao Te Ching covers the five approaches that survive the book's density.
Common questions
Is Stephen Mitchell's Tao Te Ching a good translation?
It is not a translation and does not claim to be — the book says "version" on it. Mitchell does not read Classical Chinese and worked from existing translations plus Zen training. As English poetry it is superb; as a source text it is unreliable.
Which Tao Te Ching translation is most accurate?
D. C. Lau for the standard scholarly text, Addiss and Lombardo for accuracy that still reads well, and Red Pine if you want the Chinese commentators alongside. Our comparison covers the trade-offs.
Why do scholars criticise Stephen Mitchell?
Because passages get smoothed into serenity the Chinese lacks, political advice to rulers becomes personal spirituality, contested lines arrive fully resolved, and nothing marks which lines are Laozi and which are Mitchell. Fine in a poem, a problem in a source.
Should I still buy the Mitchell version?
If you want to read the book rather than study it, yes, and pair it with an accurate edition for about the price of a paperback. It is the version most likely to actually get read, which is not nothing for a book people abandon.
What is the difference between a version and a translation?
A translation works from the source language and answers to it. A version works from other translations and answers to the reader. Both are legitimate when labelled; only one of them settles arguments about what the text says.
Is Ursula K. Le Guin's Tao Te Ching a translation?
No, and she said so plainly — she called it a rendition, worked from Paul Carus's text and other English versions, and documented her method. It is the closest comparison to Mitchell and rather more forthcoming about what it is.