Comparison of K-Ar and ages based on U-Th-He on young basalts from the San Francisco Volcanic Field near Flagstaff, Arizona, shows that approximately one fifth of the He is retained and thus U-Th-He ages are too young. Leaching with weak acid produces larges losses of U in only a few minutes, which suggests that He is lost from surifical positions on minerals from radiation damage sites and from interconnecting crystal defects. One sample of large plagioclase phenocrysts showed excess He. It seems unlikely that the U-Th-He method can be used to obtain reliable ages on young basalts.