I would highly recommend this book to anyone interested in historical linguistics, even those with little background knowledge of the subject (although note that it does assume some basic background knowledge in rudimentary linguistic concepts like "phoneme," "syntax," or "nominative"). One of the largest benefits of the book, in my view, is the fact that Trask, who was one of the foremost experts on the Basque language, frequently uses examples from Basque to illustrate points, very useful additional perspective for those with a knowledge only of Indo-European languages. Trask still, however, uses plenty of examples from the Indo-European field, and does not neglect such famous topics as Grimm's Law or the English Vowel Shift. He provides numerous references for further and more in-depth reading for a given chapter's topic at the end of the chapter, and a brief description of each of these references.
He also, as mentioned by a previous reviewer, approaches the debate on alternatives to the comparative method much more even-handedly than some of his colleagues. Although he does dismiss Merritt Ruhlen's Proto-World proposals (offering a good deal of evidence in support), he nonetheless does his best to be more neutral for other controversies, such as the Nostratic hypothesis (the proposal linking the Indo-European, Altaic, Afro-Asiatic, Dravidian, and Caucasian language families [and sometimes others]). He even asks other linguists to approach the Nostratic hypothesis fairly and with an open mind. Although he ultimately takes a stance against most radical proposals of this type, he nonetheless gives them a much more balanced portrait, and gives enough opposing arguments to allow readers to draw their own conclusions, which, considering the controversies in the world of historical linguistics today, is particularly impressive.
Ultimately, the one significant problem, also touched on by other reviewers, is that the excercizes at the end of each chapter have no solutions, and for those who, like me, study linguistics on their own and not in an academic institution, it is often frustrating to not know if they are doing the problems "right" or not.