Be advised that this is definitely not a first book, and that most of the covered topics have simpler introductory treatments elsewhere. This book is mainly a pure semantics work, along the lines of Winskel, Gunter, Tennent, Slonneger, etc. It has no discussion of implementation techniques for any of the covered topics. For that, you're better off with Mitchell's other book, or Sethi or Friedman/Wand/Haynes. Topics covered are: axiomatic, structural operational, and denotational semantics, PCF (including the full abstraction problem), universal algebra, typed lambda calculi and their models (including imperative programs), the category-theoretic approach to domain theory, logical relations, and many chapters on type systems. Several of these topics are covered more extensively elsewhere (domains by Amadio & Curien, types by Pierce), but the coverage of each topic is fairly thorough here and scarily rigorous. There are many nice excercises at the end of every section.

There are some missing topics that I wish had been included: coinduction and material on concurrency, which are not even mentioned. You'll need to get Milner's or Sangiorgi's books for this (or deBakker + de Vink's _Control Flow Semantics_). As it is, though, the book is already nearly 850 pages long.

Overall, it's a very good textbook for a grad-level semantics course, and an excellent general reference.

The book does not lack disturbingly much from mathematical completeness, however the mathematics that it fits onto the subject is not for understanding, I think the writer himself have not understood what he has typed, either, in particular I think that the universal algebraic setting that he used in Chapter 3 has better presentations; better in rigor and understandability in libraries, e.g. by Burris and Sankappavar.
I'm not a PL expert, but if it is PL and you're working in it, please do move onto a better subject, such as astrology; at least you'll not be developing self-consistent but most of the time useless (not only in practical life) systems of definitions, propositions, theorems and corollaries.

I look forward to seeing the "Springer version" of this book; maybe more expensive, rigorous, short but complete and most important of all understandable.

(signed)Turan Yuksel