This book is inspiring!
The coverage is very broad and it can serve as a handbook for graduate students. Even the hardest bits of econometrics are explained in an easy to understand way.
My personal favourites are chapters 16, 17 and 18.
I will definetely buy a hardcover of the next edition of this econometrics masterpiece.
Summing up, it's one of my favourite books.

This book is poorly organized, badly written, and so difficult as to be virtually incomprehensible to all but the most advanced mathematics students. Even the infrequent examples skip so many steps that they are hard to follow. Do not torture yourself with this book unless it is absolutely required - and unless first you have vigorously but unsuccessfully lobbied against its use. Pair this textbook with a sub-par instructor and you will be living a nightmare. Surely there is a more user- friendly option available!

Greene's book (or tome, more accurately) is a detailed treatement of contemporary econometric practices. It assumes prior knowledge in econometrics, statistics, and linear algebra. That said, it is possible to use Greene as a primary reference for a first course in econometrics, provided the stat and math background is present (if it's not, I suppose the appendices offer an abbreviated course).

Greene uses almost exclusively matrices, which is nice because it is practical (ie, all contemporary data sets are in matrix form) but also somewhat confusing as it hides some of the intuition. Where Greene skips over a derivation in, say, summation form, a more basic book would be inclusive. It is personal preference as to whether this is good or bad for the reader. For my first econometrics course, I supplemented this with Wooldridge, but ended up using exclusively Greene by the end of the quarter (Wooldridge was too verbose).

As I have continued to take more econometrics, I have started using Greene as a reference book rather than a primary source. In this capacity, I struggle to imagine a more useful book. All theorems are developed, all results are derived (at least in matrices), and the index is comprehensive.

Again, as a first book, Greene might be too rigorous or too cursory. After a short introduction to the craft, it will become an indispensible resource.

To be fair, I've only used this book for an undergrad microeconometrics course, so take that for what it is worth. The book is difficult. Ideas are often not presented in a clear manner, and the few examples are very often of poor quality. That being said, the book is probably best used as a applied reference and not as a text book. It won't steer you wrong, but if you're looking for good examples to help guide you through the fog or for well presented proofs they really aren't there. But any time I've got a problem I go for Greene.

I used the first edition of this book as a graduate student in my econometrics class. The book is very confusing and hard to understand. It's still of no help to me now that I am a professor and writing research paper. Whenever I need a review in econometrics I use Judge et all; "The Theory and practice of econometrics." It is a better book, well written, and easy to follow. Unfortunately it is off print. We need a better book than the one by Greene. If it was for me, this book should be put where it belongs, in the garbage.