The inhibitory effect of poly(A)poly(U) on the pulmonary metastasis of B16-F10 melanoma was examined in comparison with that of poly(I,C)-L,C and poly(I)poly(C). The correlation between interferon (IFN) level and antimetastatic effect was also investigated. Intraperitoneal injection of poly(A)poly(U) (50 mg/kg) into C57BL/6 mice 24 h before intravenous inoculation of B16-F10 melanoma (1 X 10(5] caused a significant decrease (p less than 0.01) in the number of pulmonary nodules 14 days after tumor challenge. But poly(I,C)-L,C (1 or 0.2 mg/kg) and poly(I)poly(C) (5 mg/kg or 1 mg/kg) did not. From the kinetic study of IFN levels induced by polyribonucleotides, poly(I,C)-L,C showed the most potent IFN-inducing activity, followed by poly(I)poly(C) and poly(A)poly(U), in this order. Plasma IFN reached a peak at 6 h and still continued to be detected at 24 h after intraperitoneal injection of the polyribonucleotides. Against B16-F10 melanoma, the cytotoxicity of spleen cells stimulated by poly(A)poly(U) (50 mg/kg) was significantly (p less than 0.05) higher than that of spleen cells stimulated by poly(I)poly(C) (5 mg/kg) both at 12 and 24 h after intraperitoneal injection of those agents. The above results that there is no correlation between the IFN levels induced by polyribonucleotides and their antimetastatic effect. More extensive study of poly(A)poly(U) might give more fruitful results, which will give valuable information for future clinical trials of this lowly toxic promising agent.