This paper presents a method for identifying and assessing the signifi
cance of interaction patterns among various chemicals and chemical cla
sses of importance to regulatory toxicologists. To this end, efforts w
ere made to assemble and evaluate experimental data on toxicologically
significant interactions, to use this information to characterize the
consistency of toxicological interactions, and to define classes of c
ompounds that display similar toxicological interactions, The motivati
on for this effort is to be able to propose hypotheses, which can be v
alidated by experimentation, on how 2 or more chemicals will interact.