Nineteen patients with stage IV melanoma were treated in an escalating dose, phase 1 trial of a DNA plasmid vaccine pSEM. The plasmid encoded T-cell epitopes from differentiation antigens Melan-A/melanoma antigen recognized by T cells (MART)-1 and tyrosinase, encompassing amino acids 26-35 and 31-70 from Melan-A/MART-1, and 1-9 as well as 369-377 from tyrosinase. End points of the trial were safety, tolerability, and melanoma antigen-specific immunity by tetramer assay. Intralymph nodal infusions of the vaccine were given 4 times, every 2 weeks over 96 hours each to groin lymph nodes. Vaccine doses were 500, 1000, and 1500 microg of DNA per infusion. Disease evaluation was performed 8 weeks after treatment initiation. The vaccine was well tolerated, with only grade I/II toxicity observed and no dose limiting toxicity at the highest dose of 1500 microg per infusion. Immune response defined prospectively was seen in 4/19 patients, and 5/19 had evidence of preexisting immunity to Melan-A/MART-1. No immune responses to tyrosinase was seen. There was a correlation between time to progression (TTP) and Melan-A/MART-1 immunity (preexisting or induced) for all patients. There was no association between TTP and immune competence assayed by ex vivo polyclonal stimulation of peripheral blood mononuclear cells. No clinical responses were seen. DNA plasmid pSEM vaccine was well tolerated when administered intranodally by 96-hour infusion to patients with stage IV melanoma, and was immunogenic, but did not induce regression of established disease. The association of TTP with preexisting or induced Melan-A immunity supports future attempts to induce potent immunity to this antigen.