Heidegger's most popular text, Being and Time (1927), was first published in English in 1962, translated by John Macquarrie and Edward Robinson. At that time only a few of Heidegger's essays and one lecture course had been translated, and there was little precedent and scholarship on which to base all the decisions the authors had to make. Heidegger exploited the German language in subtle and idiosyncratic ways to describe his original way of thinking about philosophical themes that had gone unexamined for millenia. The translation was studied by generations of scholars and over time, as it was better understood, better ways of expressing Heidegger in English were proposed.

In 1996 a new translation, by Joan Stambaugh, was published. It was hailed as a major step forward for understanding Heidegger, immediately became the new standard, and the earlier translation stopped being printed. However, after some time, sales of the original translation began to pick up, a paperback version was released, and then withdrawn when the hardback was reprinted, and it returned to its former spot as the required text for university classes. Today, ten years later, the Macquarrie and Robinson translation (M&R) outsells Stambaugh's.

What happened? Two things. First, the earlier translation was and remains the standard. In the vast secondary literature on Heidegger, references are usually to its page numbers. If you are going to study Heidegger, you will need a copy of M&R to look things up in. And second, the Stambaugh translation is simply not that much better. At least not sufficiently better to replace the existing standard.

Theodore Kisiel, probably the top authority on how B&T was written, had this to say in The new translation of Sein und Zeit: A grammatological lexicographer’s commentary,

The full translation of the book was completed in the late seventies and is thereby dated, having circulated in manuscript form for well over a decade. It therefore cannot claim to have taken the full measure of “the insights of the past thirty years of Heidegger scholarship in English,” despite the valiant latter-day efforts of the SUNY editors to update it. The “new” translation of key notions, which in some cases are left unexplained and unjustified, leaving one with the impression of unilateral willfulness, should by and large not be made to “serve as the standard for Heidegger studies to come.”...The translator’s hope to “remedy some of the infelicities and errors of the previous translation” is only occasionally and imperfectly met, in some cases in fact repeating its errors verbatim, indicating the degree to which the translator herself is obligated to the “first cut” made by M&R to decipher Heidegger’s idiosyncratic syntax and style in rendering this ground-breaking book into English. Not that the new translation followed the old in any diligent and thoroughgoing fashion. Would that this were so. Repeated comparison of the two together, against the original Niemeyer edition, is in fact one good way of uncovering the plethora of minor errors and omissions that have somehow been “left” in or “crept” into the new translation and, along the way, of acquiring a profound appreciation for the scholarly accuracy of M&R’s rendition.
...
In the many decisions involved in retranslating a great work, Heidegger’s Swabian maxim of advice drawn from the idiom of habitual everyday concerns, “lassen es bewenden,” suggests a pragmatic English equivalent: “If it ain’t broke, don’t fix it.”

Thomas Sheehan in his essay “Let a hundred translations bloom!” A modest proposal about Being and Time notes:

Many of us first met this translation some twenty years ago in its then typed format--690 double-space pages replete with hundreds of handwritten corrections. Now two decades later, a glance at that earlier manuscript reveals that little has changed in the intervening years: The published book is virtually identical to the earliest typed manuscript. So too, the Introduction here (JS 1-35) is the same one that appeared in Basic Writings (1977, 41-89), with only minor orthographical changes.

Read both essays for detailed critiques of the translation, but here's one of my pet peeves.

The Greek gets badly mangled. To begin with, the Greek is printed in hodgepodge fashion. The epigram from the Sophist is set in Greek characters, but in the body of the text all Greek words are transliterated. What is worse, the transliterated Greek is riddled with orthographical errors, misspellings, and inconsistencies. This may seem trivial; but it is just another example of how the editorial quality of JS lags far behind MR. The simple solution would have been to set all the Greek (not just the epigram) in Greek characters. Surely if readers do not understand a text printed in Greek, there is little chance they will understand it any better when transliterated.

If M&R could do the Greek correctly four decades ago, there's no excuse for botching at a time when the most popular word processor could handle Greek characters.

So, here's hoping that we'll some day be thankful for a decent translation of the definitive Gesamtausgabe edition of Sein und Zeit.  ¶ 9:01 AM