In a series of 6 experiments with CD8F1 mice with spontaneous mammary adenocarcinomas Sugiura noted by macrovisual observation with some histology an overall average of 21% of mice with lung metastases when treated with 1,000--2,000 mg/kg/day of amygdalin compared with 90% of the control mice. The significance attributed to those early observations is seriously challenged by the negative findings of 3 independent investigators, by 2 out of 3 negative cooperative experiments in which Sugiura participated, and particularly by the blind experiment in which he and others under blind readings found no anticancer activity. Treatment of Swiss albino mice showed no destructive effect upon their spontaneous mammary adenocarcinomas. Of the treated mice, 22% were found by macrovisual observation to have lung metastases while 91% were noted among the controls. The results are subject to questions raised in the discussion. Amygdalin at 2,000 mg/kg/day was ineffective both in treating and preventing the development of spontaneous leukemia in AKR mice. At 1,000 mg/kg/day it was not found effective in preventing or significantly delaying the development of spontaneous mammary tumors in CD8F1 mice. In summary, we do not have evidence to support taking amygdalin to clinical trial, although other considerations may require that one be conducted.