Objectives The purpose was to clarify the potential association between fish intake and risk of heart failure. Methods We searched relevant studies in PubMed database through January 2012 and manual review of references. Five independent prospective cohort studies involving 5273 cases and 144,917 participants were included. The summary relative risk estimates (SRRE) based on the highest compared with the lowest category of fish consumption were estimated by variance-based meta-analysis. In addition, we performed sensitivity and dose-response analyses to examine the association. Results the result is an absence of an association fish intake and heart failure (SRRE = 1.00, 95% CI: 0.81-1.24). However, fried fish intake positively associated with heart failure (SRRE = 1.40, 95% CI: 1.22-1.61). In addition, dose-response analysis of fried fish suggested that each increment of 6 times fried fish per month corresponded to a 37% increase of heart failure rate (RR = 1.37, 95% CI = 1.20-1.56). Conclusions Heterogeneity is a major concern in meta-analysis. A remarkable heterogeneity was observed across the including studies. This might be caused by variability among the study populations, follow-up period, analytic methodology, dietary assessment method and adjustment for confounding factors. Based on our subgroups by geographical region, sex, method of fish cooking, little heterogeneity was observed among studies conducted in Europe and studies that assessed fried fish individually. We were not able to other subgroup analysis due to the very limited data. However, the results of sensitivity analyses were similar and robust, indicating that no any single study considerably influenced the overall risk estimate between fish intake and heart failure. In addition, we observed no evidence of publication bias in our meta-analysis based on Egger’s test and Begg’s rank correction. Our results compare favorably with most of studies included in our analysis, in which they reported that subjects consuming fish were not associated with heart failure risk, whereas consumption of fried fish was associated with 40% higher risk for incident HF. Dose-response analysis of fried fish suggested that 37% increased risk of heart failure were caused by incremental increase of consuming on average 6 times fried fish per month. The underlying mechanism involved in the association between fried fish consumption and heart failure is uncertain. One plausible reason is that the net effect of benefit versus risk of fried fish consumption may be detrimental (29). Although method of frying does not decrease the absolute n-3 fatty acid level, frying adds other fatty acids from the frying oil and the procedure of cooking at high temperature may add oxidation products, partially hydrogenated oils and transfatty acids (30-32).These products may cause the HF risk increasing. In addition, the association between fried fish intake and heart failure risk appeared partly related to higher risk clinical and lifestyle factors. The higher fried fish consumption was tightly correlated with lower fiber and higher fruit/vegetable intake (15, 19).Further, it could cause higher prevalence of diabetes, atrial fibrillation, CAD, higher systolic blood pressure, higher body mass index, higher prevalence of smoking and higher energy intake (15).