Four commercially available arthritogenic adjuvants ('modified Perrigens') were compared with a Mycobacterium butyricum-mineral oil adjuvant for their ability to induce arthritis in male Holtzman rats. The induced arthritides were also compared with respect to their response to treatment with phenylbutazone and cyclophosphamide. Trivial differences were noted among the various adjuvants in their capability to induce the disease, the severity of the disease, or the response of the disease to therapeutic (phenylbutazone) or prophylactic (cyclophosphamide) administration of drugs. It was concluded that any one of the adjuvants was probably suitable for inducing an arthritis in rats for the testing of drugs and that perhaps other variables in the assay are more important in explaining inter-laboratory variation.