Investigating the immunotoxic potential of candidate drugs as part of a preclinical safety evaluation poses several problems. These include the need for practical, validated tests, the difficulty in establishing the toxicologic significance of positive findings, and a poor understanding of the predictive value such findings hold for drug effects in man. A key component of this investigation is the toxicologic profile generated through preclinical toxicity and safety studies. As this "routine" assessment becomes increasingly comprehensive and sophisticated, most toxicologically significant drug-associated effects are revealed. Such findings may serve as "triggers" for investigating possible immune mechanisms. Decisions to test specifically for immunotoxicity may also be influenced by the molecular structure and pharmacologic profile of the compound, as well as the intended use of the drug. Examples of such indications and follow-up studies are discussed in this review. We are presently poorly equipped to effectively screen drugs indiscriminately for an immunotoxic potential. We are better prepared, however, to investigate whether a drug-associated change is due to an adverse effect on the immune system. This problem-oriented approach to immunotoxicology challenges us as diagnosticians and immunopathologists, and requires a close working relationship among the toxicologic pathologist, the basic immunologist, the immunopharmacologist, and the clinician.