International rates of breast cancer for females aged 40-44 years (the “early” rate) and for females aged 65-69 years (the “late” rate) were positively correlated with sugar and fat intakes. The correlation explained three-quarters of the variation in the late rate, for 22 countries, but only half of the variation in the early rate. The late rate was, further, positively correlated with estimates of the percentage of nulliparous women (9 populations) and, together with terms for sugar and fat intakes, the multiple regression explained 90% of the variation. Early registration rates (13 populations) were positively correlated with blood group A which appeared, from the multiple regression equation, to contribute more than twice the amount to the early rate than did sugar and fat intakes. The contribution of blood group A to the late rate appeared to be only one-third of that for sugar and fat intakes.