Preclinical data indicate that the combination of retinoids and interferons have synergistic antiproliferative and differentiating effects in some hematologic and solid tumor models. These observations have led to clinical trials in which 13-cis-retinoic acid (13cRA) 1 mg/kg/day was combined with interferon alpha-2a (IFN alpha) 3 or 6 x 10(6) U/day. The first two such trials produced exciting results: 50% response rate in patients with previously untreated stages IB-IVA cervix cancer and 68% in patients with advanced squamous cell skin cancer. These data led to a number of additional trials of the combination, but the high response rates seen in the initial cervix and skin trials have not been duplicated in the other squamous tumors tested (head and neck, lung, pretreated cervix). In addition, trials in two nonsquamous histologies were negative (lung and melanoma). However, the regimen was not always studied in an optimal population of previously untreated patients and the negative results in pretreated cervix patients point to the relevance of this consideration. Nevertheless, the observation that the combination of 13cRA and IFN alpha (both of which bind to specific receptors and change gene expression) is able to induce regression in advanced tumors, must be regarded as highly important. Key questions to be addressed include an understanding of the biologic mechanism of specific tumor sensitivity (why some squamous tumors and not others?), and mechanisms of resistance in sensitive tumor types (e.g. cervix). Such data may lead to trials targeted to tumor types with defined biologic features having a high liklihood of clinical benefit. In the meantime, studies integrating this combination with other active treatment modalities such as radiation is warranted in cervix and skin carcinomas.