OBJECTIVES: Use of meta-regression to explain heterogeneity between the studies evaluating relationship between the level of glycated hemoglobin (HbA1c) and the risk of development of retinopathy in patients with type 2 diabetes mellitus (DMT2). In meta-regression the results of the studies under consideration are treated as points to be analyzed by weighted regression. If the results obtained in a model without covariates indicate heterogeneity of studies, then inclusion of subsequent covariates into the model can make it possible to check if specific variables explain heterogeneity. The amount of heterogeneity may be expressed as between-study variance. If between-study variance decreases after inclusion of a specific covariate into the model, it means that heterogeneity between the studies may be explained by that covariate. METHODS: The following covariates were considered: observation period, type of the study, sample size, HbA1c level, age, duration of DMT2, BMI, cholesterol level, arterial blood pressure. Due to a scarce number of studies available between-study variance was estimated using the REML method. The value of between-study variance evaluated using a meta-regression model for retinopathy in DMT2 without covariates was 0.4547. RESULTS: Inclusion of the difference between the intervention group and the control group with respect to the HbA1c level resulted in decrease to 0.1572. Therefore the covariate under consideration explains the heterogeneity in 65.4%. The obtained value of directional coefficient was statistically significant and equal to 0.5824, which means that increase of difference between the groups with HbA1c level by 1 unit results in increase of the logarithm of relative risk by 0.5824, reflecting increase the risk by 1.79 times. CONCLUSION: It was demonstrated using meta-regression that the difference with respect to the HbA1c level is the cause of heterogeneity between the studies. Relative risk of development of retinopathy increases with increase of the difference in HbA1c level.