Objective To estimate the infertility rates and to examine factors that contribute to the variations in infertility rates among studies.Methods Wanfang,CQVIP,CNKI,and PubMed Database were searched for infertility relevant population-based prevalence studies between 1980 and 2012.Meta-analysis and meta-regression were used to calculate the pooled size of effect and to identify the sources of variation.Results A total of 27 articles using data from regional surveys were included in the review.Results showed that the combined prevalence was 5.7％ (95％CI:5.3％-6.1％).Combined prevalence rates of 1-year and 2-year infertility in newly married couples were 12.5％ (95％CI:9.5％-15.4％) and 6.6％ (95％CI:4.9％-8.4％),respectively,while were respectively 6.4％ (95％CI:4.2％-8.6％) and 3.0％ (95％CI:2.6％-3.3％),in women of reproductive age.Results from Meta-regression showed that the prevalence of 2-year infertility was lower than that of 1-year while infertility in regions was seen higher under the poor socioeconomic condition than in the better regions.Infertility in studies using cohort or prospective studies was seen to be higher than in other studies.Conclusion Owing to the inconsistency in determining the numerator and denominator that were used to calculate the prevalence of infertility as well as the inconsistency in criteria used to define the infertility,an agreed definition on infertility needs to be followed,in order to facilitate the comparison among studies.