The establishing of safe levels of mycotoxins to date has been a legal rather than scientific exercise. This has resulted in levels which have varied in response to economic and political pressures. The data base for rationally determining safe levels is very small. This has resulted in subjective evaluations of the worth of different studies in attempts to deduce safe levels from experiments designed to demonstrate effects, and in assumed safe levels which vary from field experiences. Using physiological parameters other than growth as criteria of safety, known deleterious interactions of mycotoxins with other factors, and statistical corrections for inadequate numbers of animals tested, permit better agreement between safe levels determined from laboratory data and from field data. However, the number of animals required makes impractical the laboratory determination of truly safe levels. Well-conceived and executed epidemiological studies coupled with laboratory studies designed to elaborate underlying principles appear to be the best approach to determining safe levels of mycotoxins. Until safe levels are based on sound animal experimentation, the prudent person would assume there is no truly safe level and that increasing levels of mycotoxins carry increasing risk.