The level of risk involved with human exposure to chemicals isusually calculated by looking at human epidemiological data and by performing toxicological tests on animals or on bacteria and animal cells. "Epidemiological data provide the best information, but when they are unavailable we have to extrapolate possible effects from animal data," says Arthur L. Craigmill, a Cooperative Extension toxicologist in the environmental toxicology department at UC Davis. "It's important to remember, however, that there are numerous sources of uncertainty involved with predicting human risk from animal data." For example, Craigmill points out that the high doses needed to produce effects in animals are often orders of magnitude greater than those to which humans would be exposed. In a review talk, he will identify these and other sources of uncertainty when he discusses the toxicological tests used in exposure assessment. His talk will be presented Wednesday, April 8, at 1:15 p.m. in the Hilton Hotel.