The science of epidemiology is not easy. It is not a matter of counting heads, counting disease, and matching against a suspect variable. To his great good credit, Gough, a former Director of Special Projects for the Office of Technology Assessment, has tackled one of the more emotional battlegrounds of the decade: the risk, be it cancer, heart disease, neurological damage or birth defects, of exposure to Agent Orange. The contaminant of Agent Orange was dioxin. As long ago as the mid-50's a German scientist noted that factory workers exposed to high dioxin levels developed cloracne, a pimply skin condition that could cause scarring and even darkening of the skin. Laboratory tests soon showed that dioxin was extremely toxic to rats, causing liver cancer and miscarriages. Such data set the background for many studies following the Vietnam War, when class-action suits were initiated by veterans complaining of fatigue, impotency, birth defects in offspring, and other ailments, including cancer. Some settlements followed, not so much because the case was proven but more out of sympathy and a desire to end the matter. There followed other incidents, the most celebrated being the chemical explosion in Seveso, Italy, and the contamination of Times Beach, Missouri, where the thickened heavily contaminated sludge from ""still bottoms"" had been used to keep the dust down on rural roads and led to the deaths of animals and to human sickness. In Seveso, children and adults, especially in the zones most severely affected by the cloud of dioxin, laden waste, developed cloracne and other symptoms. Years later, however, Seveso's citizens seem to have suffered little permanent damage; the Environmental Protection Agency has minimized the cancer threat to residents of Times Beach. All production of dioxin-contaminable products has now ceased in the US, while elsewhere manufacturing standards limit the level of dioxin to neglibile amounts. Gough concludes, after an exhausting case-by-case study, that Agent Orange has been examined and has essentially been found wanting. It is possible that a cancer death may occur as a result of exposure, but given the prevalence of cancer in this country one could never statistically blame dioxin. He recounts the numerous Vietnam veteran studies pointing to the inherent problems of doing retrospective analyses (who can truly say where all the men of a given battalion were in relation to an area sprayed?). The ""Ranch Hands""--the men who actually serviced the planes and did the spraying--have been followed and show clean bills of health. On the other hand, there are serious cases of factory workers plagued by cloracne and neurological damage after heavy exposure--still they are not succumbing to mortal disease. Add to that the many uncertainties in case ascertainments, memories, and the psychological damage experienced by many veterans and you get a glimmer of the difficulties. Stick to this book if you want insights into risk assessment by a levelheaded clear expositor. Interestingly, Gough has now moved on to become Director of the Risk Science Institute in Washington. It is comforting to have a clear head in such a place.