Behavioral teratology/toxicology: how do we know what we know?

Knowledge about human behavioral teratology/toxicology must necessarily be limited by the methods that can ethically be used. Research in the field is an enterprise calling for tolerance for uncertainty if not chaos. Among the main points of this paper are: (a) Kaufman's criticisms of research on low blood-lead level (BLL) and children's IQ are generally valid and apply to virtually all human natural-groups research; (b) Relative to some contexts, research on low BLL on children's IQ is exemplary; (c) Relative to other contexts, the conclusion that low BLL have linear effects is well supported; (d) Owing to necessary design limitations, all natural-groups studies have shortcomings; (e) Inference of causality is uncertain under any conditions; and (f) Some authors readily leap from molehills of data to mountains of causal conclusions.