A new rapid assay has been developed for measurement of the binding of [3H]retinoic acid to cellular retinoic acid-binding protein. The assay, which uses activated charcoal for the separation of bound from unbound retinoic acid, was used to determine the concentration required to inhibit the binding of [3H]retinoic acid to cellular retinoic acid-binding protein by 50% for 18 retinoids with free carboxylic acid groups. Partially purified cellular retinoic acid-binding proteins isolated from rat testes and carcinogen-induced rat mammary tumors were used for these determinations. The following parameters were also determined for some or all of the retinoids: hypervitaminosis A doses; activity against carcinogen-induced mouse skin papillomas; inhibition of growth of a rat chondrosarcoma; inhibition of growth of 3T6 cells; and differentiation of the embryonal carcinoma cell line PCC4.azalR. While all retinoids that are potent in these biological test systems bind tightly to cellular retinoic acid-binding protein, the converse is not true. The lack of a consistent quantitative correlation between 50% inhibitory concentration and biological activity is probably due to insufficient concentrations of the retinoid in the target tissue or cell, which is a consequence of factors such as absorbability, metabolism, tissue distribution, and pharmacokinetics. © 1980, American Association for Cancer Research. All rights reserved.