AIM: Diabetes mellitus affects the pharmacokinetics of cytochrome P450 3A4/5 (CYP3A4/5) substrates. We evaluated the relationship between haemoglobin A1c (HbA1c) levels and the pharmacokinetics of controlled-release tacrolimus. METHODS: This retrospective observational cohort study included kidney transplant recipients (>18 years) receiving controlled-release tacrolimus orally. CYP3A5 genotypes were categorized as expressers (*1/*1 or *1/*3) and non-expressers (*3/*3). Multiple linear regression analysis determined the predictors for trough concentration/dose-normalized by body weight (C/D) ratio of tacrolimus at 7 days, 6 months and 12 months after administration. Correlations between the C/D ratio and HbA1c levels at baseline, 6 and 12 months after tacrolimus initiation were evaluated with Bonferroni correction. RESULTS: Out of 42 patients (CYP3A5 expressers, n=56, and non-expressers, n=83), the multiple linear regression analysis showed that the C/D ratio on Day 7 was marginally higher in CYP3A5 non-expressers than in CYP3A5 expressers. Factors affecting the elevation of tacrolimus C/D ratio after 6 and 12 months of treatment were male sex and CYP3A5 non-expressers and increased HbA1c levels and CYP3A5 non-expressers, respectively. The C/D ratio and HbA1c levels after 12 months was positively correlated in CYP3A5 non-expressers (y=54.6x+194.6, R=0.63, P=0.004, Bonferroni correction). Furthermore, intra-individual changes in the C/D ratio and HbA1c levels from 6 to 12 months were nearly correlated (y=54.5x+20.2, R=0.41, P=0.036, Bonferroni correction). CONCLUSION: HbA1c and CYP3A5 genotypes might be considered to understand the inter- and intra-individual variability in blood tacrolimus concentrations after 6 months post-kidney transplantation.