The following review is from IM John Watson, originally published on The Week In Chess website
A fairly advanced instructional classic is Vladimir Vukovic's The Art of Attack, originally written in 1965, and rather enthusiastically revised and corrected by John Nunn in this new edition. Cadogan/Everyman has done the chess world a favor by reprinting this book, which is still cited more than any other by players asked about their favorite book on the attack. Vukovic classifies attacks and attacking principles into comprehensible categories, and has the art of explaining the key features of an attack in a way that the student will not forget. His analysis is both deep and original; it led to revised opinions of some of the most famous games in history. The reader can review this book more-or-less casually and learn a lot just from it's diagrammed examples and prose; but he or she can also spend hours immersed in the more complex notes Vukovic provides, and thus learn even more.
I have to admit that John Nunn's numerous corrections of Vukovic's analysis left me a little uncomfortable at first. Nunn says of his various "analytical footnotes": "I hope readers will not form the opinion that Vukovic's analysis was especially unsound--this is certainly not the case." But in fact, the corrections (in themselves flawless, as far as I can see) are so ubiquitous, and so often obvious in hindsight, that one starts to wonder (a) whose book this is; and (b) how a book on the attack can have so many tactical oversights! In a way, however, this is a good lesson for us all. The chess world has changed dramatically, however much we might not like to admit it. Analysis by strong players today is in general more accurate and thorough than that by legends of the past. Of course, some of this is due to the use of analytical engines (playing programs), and Nunn has no doubt benefited greatly from their problem-spotting and confidence-raising attributes. But the increased dynamism and complexity of modern chess, along with the simple fact of heightened competitiveness, has both improved published analysis and subjected it to more stringent criticism. Furthermore, the errors in Vukovic's book to some extent undermine his attempts to philosophize and generalize about the attack (and the history of the attack). The modern attacker (exemplified by Kasparov and Shirov) ultimately relies on few or no principles, but rather emphasizes the concrete and analytical to an extent previously unknown. Of course, intuition also plays a major role, but even intuition is informed by calculation and experience more than by what used to be called "principle." Perhaps Vukovic's analysis was not "especially unsound" for it's day, but it certainly would be now. Nunn's corrections help us to focus on the beauty and complexity of Vukovic's examples, and at the same time to be skeptical of his often-dogmatic claims about both individual players as attackers and why in general an attack "must" succeed or fail. That is certainly a healthy thing if our goal is really to improve our attacking skills.
To conclude, I think that The Art of Attack in Chess would be enjoyable to players in a range from about 1500 to 2600, and useful for players from about 1500 to 2200. Beneath 1500, my guess is that getting extremely used to solving tactical problems (from one of those "1001 Combinations" sort of books) would be the appropriate preparation for a later study of Vukovic.
Published by Everyman, 352 pages