The efficacy of Human 6 IFN (HLIFN) given in a pulse fashion was determined in a phase II study. Ninety-one cancer patients were evaluated (9 myeloma, 12 breast, 14 prostate, 9 melanoma, 4 renal, 6 astrocytoma, 7 ovarian, 9 large bowel, 7 gastric, 14 head and neck). They all had advanced progressive cancer that was resistant to chemotherapy and/or radiotherapy. Patients were treated by intramuscular injection of 6 X 10(2) I.U./m2 for three consecutive days every four weeks. 84 patients were evaluable. Complete clinical response was obtained in 23 patients (4 myeloma, 2 breast, 5 prostate, 1 melanoma, 1 renal, 2 astrocytoma, 2 ovarian, 2 large bowel, 1 gastric, 3 head and neck). Partial responses were observed in 35 patients (3 myeloma, 7 breast, 6 prostate, 4 melanoma, 1 renal, 2 astrocytoma, 3 ovarian, 4 head and neck). Objective responses were related (P less than 0.01) to serum IFN level, with complete and partial responses (P less than 0.01) more commonly seen in those patients whose serum IFN levels at two hours were in the range of 1000 to 1650 I.U./ml. Side effects resulting from pulse IFN were acceptable for this group of patients and consisted of fever, transient chills, malaise and asthenia, and transient thrombocytopenia and leukocytopenia. The extent of fever was directly related (P less than 0.01) to response, and was most elevated in patients who achieved objective responses. IFN administered in a pulse fashion appears to be more effective than daily IFN and merits further evaluation.