A considerable body of work has demonstrated that phenolic polychlorinated biphenyl (PCB) metabolites, structural analogues to estradiol, bind to the soluble estrogen receptor (ER) and that hydroxy PCB-ER complexes will translocate into the nucleus and bind to ER response elements in cultured cells. Although catechol estrogens exhibit weak estrogenic activity, the catechol PCB metabolites which are structurally similar to these ER agonists have gone untested for potential estrogenicity. In the present work we have assessed the estrogenicity of this second group of PCB metabolites, the catechols. The test compounds used in the present study were chosen to elucidate the effects of chlorine and catechol position on in vitro estrogenicity. Cultured HeLa cells, transfected with the estrogen reporter gene ERET81CAT and mouse ER cDNA, were incubated with PCB catechols. The cells were harvested at 28 h posttransfection and assayed for chloramphenicol acetyl transferase (CAT) activity. The responses elicited by the PCB catechols tested fell within the range of effect measured for the catechol estrogens and phenolic PCBs, and were within the range previously reported for other "environmental estrogens" such as nonylphenol and o,p'-DDT. Maximal measured responses were achieved at concentrations approximately two to three orders of magnitude higher than that of 17-beta-estradiol, indicating that PCB catechols have estrogenic activity in vitro. The extent of chlorination and the position of the catechol (3,4 vs 2,3 substitution) were important in determining estrogenicity in the compounds tested. The 2,3-catechol showed no detectable activity in this system, while activity of the 3, 4-catechols increased with the degree of chlorination. The observed estrogenicity of PCB catechols suggests that further oxidative metabolism of estrogenic PCB phenolic metabolites would not necessarily result in lowering the total estrogenic burden of a PCB-exposed organism. The present results imply that if estrogenic activity is assigned to an individual phenol, the potential contribution of its catechol metabolites to the total estrogenic burden should also be taken into consideration.