The omega-3 fatty acids, especially long-chain eicosapentaenoic acid (20:5n-3) and docosahexaenoic (22:6n-3) contained in "fatty" fish, have consistently been shown to retard the growth of breast cancer in vitro and in animal experiments. In contrast, studies of the association between fish consumption and breast cancer risk in human populations have not consistently shown inverse associations. However, previous studies have not considered the specific types of fish consumed. Using data from a large, nationwide case-control study conducted in Sweden, we examined the association between consumption of fatty and lean fish and breast cancer risk. Odds ratios (OR) and 95% confidence intervals were computed from unconditional logistic regression models. High consumption of fish was weakly associated with reduced breast cancer risk, and the association was not statistically significant. With multivariate adjustment, the OR for women with the highest consumption (> or =3.5 servings/wk) compared with women with the lowest (virtually none) was 0.88 (95% confidence interval = 0.60-1.29, P for trend = 0.15). When type of fish was examined separately, the association was similar for fatty and lean fish.