Background: Preterm birth (PTB) has been recognized as a crucial long term risk factor for multiple non-communicable diseases. However, studies between the relationship of PTB and the risk of acute childhood leukemia have yielded inconclusive results. Therefore, we performed a meta-analysis to systematically review the current literature to investigate whether PTB is associated with increased risk of acute childhood leukemia. Methods: Three electronic databases (PubMed, Web of Science, and EMBASE) were searched up to December 1st, 2015. Relevant studies reporting the association between PTB and subsequent risk of acute childhood leukemia were included for further evaluation. Statistical analysis was performed using Revmen 5.3 and Stata 10.0. Results: A total of 12 studies for acute childhood leukemia, eight studies for acute lymphoblastic leukemia (ALL), and seven studies for acute myeloid leukemia (AML) were included in the current meta-analyses. We calculated pooled odds ratio (OR) and 95% confidence interval (CI) to evaluate the relationship between PTB and acute childhood leukemia as well as its two subtypes: ALL and AML. Our results suggested that PTB was significantly associated with increased risk of acute childhood leukemia (OR = 1.09, 95% CI = 1.02-1.17, P = 0.01) and AML (OR = 1.42, 95% CI = 1.21-1.67, P < 0.01). However, PTB was not significantly associated with an increased risk of ALL (OR = 1.04, 95% CI = 0.96-1.13, P = 0.29). Conclusion: Our data showed that PTB increased the risk of AML. Further studies are required to explore causality and dissect the biological mechanisms involved.