We carried out this meta-analysis to explore the influence of paraoxonase 1 activity on the susceptibility of diabetes mellitus (DM), diabetic macroangiopathy and diabetic microangiopathy. Relevant studies were identified from PubMed, Web of Science and CNKI without language limitation, following the inclusion and exclusion criteria. Statistical analyses were implemented with the STATA 12.0 statistical software. Thirty-six case-control studies were included in the meta-analyses, in which 35 for the association between paraoxonase 1 activity and DM risk, 8 for diabetic macroangiopathy and 7 for diabetic microangiopathy. Paraoxonase 1 activity was significantly associated with the susceptibility of DM in pooled population (SMD = -1.37, 95% CI = -1.79 ∼ -0.96, P = .000), and Asians (SMD = -2.00, 95% CI = -2.56 ∼ -1.44, P = .000), but not in non-Asians (SMD = -0.44, 95% CI = -0.91 ∼ 0.03, P = .069). However, marked heterogeneity was existed (I  = 98.10%, P = .000) and subgroup analyses failed to investigate the sources of heterogeneity. Then, meta-regression was performed and found that ethnicity could explain the observed between-study heterogeneity (P = .002). Meanwhile, significant associations were found between paraoxonase 1 activity and diabetic macroangiopathy (SMD = -1.06, 95% CI = -1.63 ∼ -0.48, P = .000) and diabetic microangiopathy (SMD = -0.72, 95% CI = -1.32 ∼ -0.13, P = .018). In conclusion, paraoxonase 1 activity plays important roles in the risk of DM, diabetic macroangiopathy and microangiopathy with ethnicity differences. Further studies with large sample and well design are needed to confirm these results.