Background: The issue of whether saturated fats and trans fats are superior predictors of all-cause death and cardiovascular disease than n-3 polyunsaturated fatty acids, such as eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), remains a matter of contention. Furthermore, few studies have examined the relationship between fatty acids and the outcomes of cardiovascular disease (CVD) in Asian populations. The aim of this study was to compare the effectiveness of various plasma fatty acids as predictors for all-cause death and CVD events in an ethnic Chinese population. Methods: This study assembled a community-based prospective cohort, comprising 1833 participants (60.6 +/- 10.5 yrs, 44.5% women) who underwent a comprehensive evaluation of fatty acids in blood using gas chromatography. None of the subjects had a history of CVD at the time of recruitment. Results: A total of 568 individuals died and 275 individuals developed CVD during the follow-up period (median of 9.6 years; interquartile range of 8.9-10.5 years). Following adjustment for established cardiovascular risk factors, the relative risk of all-cause death in the highest quartile, compared with the lowest quartile, was 1.33 for saturated fats (95% confidence interval [CI], 1.01-1.75, test for trend, P = 0.015), 1.71 for trans fats (95% CI, 1.27-2.31, test for trend, P = 0.0003), 0.77 for EPA (95% CI, 0.59 -1.00, test for trend, P = 0.048), and 0.89 for DHA (95% CI, 0.68-1.18, test for trend, P = 0.354). Similar patterns were observed for CVD events. Trans fats presented the largest area under the receiver operator characteristic curve (0.740, 95% CI, 0.716-0.766) for the prediction of all-cause death. A mutually adjusted two-marker model indicated that saturated fats and trans fats were significant predictors of all-cause death and CVD; however, the other fatty acids were not. In addition, trans fats presented the greatest improvement in net reclassification for all-cause death (7.7%, P = 0.003), followed by EPA (3.8%, P = 0.033). Saturated fats presented the greatest improvement in net reclassification for CVD events (5.6%, P = 0.039). Conclusions: Our data provides strong evidence to support that plasma saturated fats and trans fats can predict all-cause death and CVD more effectively than other fatty acid markers. (C) 2013 Elsevier Ireland Ltd. All rights reserved.