Candidate-gene association studies that examined the association between polymorphisms of the inhibin alpha gene (G769A, C16T and A124G) and premature ovarian failure (POF) have reported contradictory results. Thus, a meta-analysis of these studies was carried out. The random effects odds ratio (OR) with the corresponding 95% confidence interval and the heterogeneity among studies were estimated. Existence of potential bias and consistency of effect sizes across ethnicities were explored. Cumulative meta-analysis was also performed. The studies provided 1030/1660, 936/1398 and 938/1446 cases/controls for G769A, C16T and A124G polymorphisms, respectively. The meta-analysis showed significant heterogeneity among the studies (P(Q) = 0.01, I(2) = 74%) and lack of evidence that carriers of the G769A variant confer risk of POF: OR = 1.38 (0.48-3.94). Asian Indians (only two studies) produced significant association [OR = 8.10 (1.27-51.6)]. Regarding C16T and A124G polymorphisms, 16T and 124G alleles were not associated with POF: OR = 0.94 (0.76-1.16) and OR = 0.98 (0.86-1.11), respectively. The cumulative meta-analysis for G769A and C16T polymorphisms showed a trend in time towards to non-significance for both polymorphisms. Cumulative meta-analysis indicated that more evidence is needed to draw safer conclusions regarding the effect sizes. There was no differential magnitude of effect in large versus small studies. In conclusion, there is no evidence of association between the studied polymorphisms and POF.