This book has legs! I read it first in graduate school in 1969. I was impressed with the argument then, and still appreciate its power now. He identifies a key problem as (page 12) "identifying and achieving the conditions of peace. . ." He notes that, over time, three separate views have dominated discourse on the causes of war (and how to achieve peace): (a) human nature is the root cause; (b) the structure of states is the key factor; (c) the international system itself is the major variable.
The book proceeds in a linear fashion. First, he examines the variety of arguments locating the cause of war in human nature. However, he also notes that to link human nature to war is not easily done (there is, of course, much debate over exactly what human nature is--or even if there is such a given nature), and that political matters must be taken into account. As he considers the contributions of the behavioral sciences, he notes that (page 79) "The more fully behavioral scientists take account of politics, the more sensible and the more modest their efforts to contribute to peace become."
The second level of analysis is the structure of states themselves. He notes that some have argued that if the state had a proper structure, then peace would result. He considers, for instance, liberal theorists of the 19th century who made that point. One problem: While trying to create more liberal states, what about those illiberal ones who may engage in conflict? What then? The structure of the state won't prevent self-defense. Indeed, some liberals, like Thomas Paine, wanted to use force to democratize the world.
The final level of analysis is the structure of the international system itself. The main point here is that that system can be termed "anarchy." There is no central force to prevent outbreaks of violence. So, violence will occur. Interestingly, he begins the chapter on international anarchy with a quotation from Cicero (page 159) "For what can be done against force without force." States need to protect themselves when there is no mechanism to maintain peace; they will act in their national interest when threatened. The end result is the possibility of war whenever a country might be threatened. In Waltz' words (page 227): "According to the third image, there is a constant possibility of war in a world in which there are two or more states each seeking to promote a set of interests and having no agency upon which they can rely for protection."
In short, all three levels (images) must be understood. None is irrelevant. But the key to understanding war is the state of international anarchy. The book holds up well over time. It still presents a useful message, albeit from the hard-nosed realist position. Neocons won't like the argument that changing the structure of states won't make a lot of difference as long as there is international anarchy. Anyhow, for those interested in a fairly hard-headed analysis, this book still serves a useful purpose.