Iso H (Department of Epidemiology and Mass Examination, The Center of Adult Diseases, Osaka, 1-3-3 Nakamichi, Higashinariku Osaka 537, Japan), Sato S, Folsom A R, Shimamoto T, Terao A, Munger R G, Kitamura A, Konishi M, lida M and Komachi Y. Serum fatty acids and fish intake in rural Japanese, urban Japanese, Japanese American and Caucasian American men. International Journal of Epidemiology 1989, 18: 374–381. To examine the relationship of fish intake with serum fatty acids cross-culturally, we surveyed a total of 136 men aged 34 to 55 years in four different populations: rural Japanese (RJ), urban Japanese (UJ), Japanese Americans (JA), and Caucasian Americans (CA). Mean levels of estimated total fish intake per day were 124.9 g in RJ, 70.8 g in UJ,45.7 g in JA, and 32.3 g in CA. The percentage of total serum fatty acids contributed by omega-3 polyunsaturated fatty acids in these populations was 11.8% in RJ, 9.0% in UJ, 3.4% in JA and 2.5% in CA. Means of omega-3 fatty acids and intake of fish were correlated at the population level; this relationship was strongest when intake of dark-meat fish was considered (n = 4, r = 0.979, p = 0.02). Within each population, except for UJ, individual omega-3 fatty acid levels and dark-meat fish intake were significantly associated (p<0.05). A 20 g increase in dark-meat fish consumption was associated with an estimated relative increase in omega-3 fatty acids content of serum by 0.76% in RJ, 0.75% in UJ, 0.64% in JA, and 0.22% in CA. The association between fish intake and serum omega-3 fatty acids at the individual level was not explained by other coronary risk factors. Mortality from coronary heart disease is much lower in Japan than in the US. Population differences in fish intake and serum omega-3 fatty acid levels may contribute to the population difference in the risk of coronary heart disease.