For active specific immunotherapy, autologous tumor cells grown in vitro may be a more appropriate source of tumor antigen than allogeneic tumor cell lines, autologous tumor cell suspensions, or purified/synthetic tumor antigen. A major limitation to this approach, however, has been the ability to reliably grow tumor cells from a high percentage of fresh tumor samples. We have harvested fresh tumors and attempted to establish short-term cultures of tumor cells to obtain 10(8) cells which could subsequently be used in autologous tumor cell vaccine programs. Fresh tumors were mechanically processed to initiate primary cultures in RPMI-1640 containing 1 mM sodium pyruvate, 2 mM glutamine, 10 mM N-(2-hydroxyethyl)piperazine-N'-(2-ethanesulfonic acid), 15% fetal bovine serum, and antibiotics, incubated at 37 degrees C in 5% CO2. We were successful in growing 87 of 142 [61%, (95% confidence limits [55-68%]) of all tumors] including 39 of 58 (67%) melanomas, 10 of 10 (100%) renal cell carcinomas, 14 of 14 (100%) sarcomas, and 23 of 54 (43%) various adenocarcinomas. Success rates were not significantly higher in tumors obtained locally that were processed within 4 h of surgery, 51 of 78 (65%) versus 36 of 64 (56%) of those from farther away with longer delays in processing (p = 0.276). Of the 87 tumor cell lines established, 51 have been expanded for use in autologous tumor cell vaccine programs, and 40 have been used in the treatment of patients. We conclude that autologous tumors can be grown in vitro with sufficient reliability to make an autologous tumor cell line vaccine trial feasible.