The purpose of this study is to clarify the association between Zn levels and myocardial infarction (MI) using a meta-analysis approach. We searched articles in the PubMed, OVID, and ScienceDirect published as of November 2014. Thirteen eligible articles with 2886 subjects from 41 case-control studies were identified. Overall, pooled analysis indicated that subjects with MI had lower Zn levels than healthy controls (standardized mean difference (SMD) = -1.848, 95 % confidence interval (CI) = [-2.365, -1.331]). Further subgroup analysis found that subjects with MI had lower Zn levels than healthy controls in serum (SMD = -1.764, 95 % CI = [-2.417, -1.112]) and hair (SMD = -3.326, 95 % CI = [-4.616, -2.036]), but not in toenail (SMD = -0.396, 95 % CI = [-1.114, 0.322]). The subgroup analysis stratified by type of Zn measurement found a similar pattern in inductively coupled plasma-atomic absorption spectrometry (ICP-AAS) (SMD = -2.442, 95 % CI = [-3.092, -1.753]), but not in neutron activation analysis (NAA) (SMD = -0.449, 95 % CI = [-1.127, 0.230]). Lower Zn levels in MI patients were found both in male (SMD = -3.350, 95 % CI = [-4.531, -2.169]) and female (SMD = -2.681, 95 % CI = [-3.440, -1.922]). And the difference of Zn levels according to MI in Asia (SMD = -2.555, 95 % CI = [-3.267, -1.844]) was greater to that among the population in Europe (SMD = -0.745, 95 % CI = [-1.386, -0.104]), but no difference was found in Oceania (SMD = -0.255, 95 % CI = [-0.600, 0.089]). In conclusion, this meta-analysis indicates that there is a significant association between Zn deficiency and MI. We suggest that a community-based, long-term observation in a cohort design should be performed to obtain better understanding of causal relationships between Zn and MI, through measuring hair Zn at baseline to investigate whether the highest zinc category versus lowest was associated with MI risk.