Objective To investigate the relationship between non-alcoholic fatty liver disease (NAFLD) defined by ZJU index and serum 25-hydroxyvitamin D levels in patients with type 2 diabetes mellitus (T2DM). Methods The clinical data of 917 T2DM patients diagnosed in Department of Endocrinology, First Hospital of Lanzhou University from Apr. 2016 to Oct. 2020 were collected. The patients were divided into non-NAFLD group (ZJU index＜32.0), uncertain group (ZJU index 32.0-38.0) and NAFLD group (ZJU index＞38.0), and the serum 25-hydroxyvitamin D levels of the 3 groups were compared. The association between 25-hydroxyvitamin D levels and NAFLD defined by ZJU index was analyzed by generalized linear model. Results There were 570 NAFLD patients (62.16%) defined by ZJU index in 917 T2DM patients. The ZJU index defined 336 (81.55%) patients as NAFLD in 412 T2DM patients who were diagnosed with fatty liver by ultrasound. The prevalence rates of NAFLD in T2DM patients with hypertension, hyperlipidemia, overweight or obesity were significantly higher than those in T2DM patients without hypertension, hyperlipidemia or with normal weight (64.92% ［396/610］ vs 56.68% ［174/307］, 68.18% ［480/704］ vs 42.25% ［90/213］, and 76.83% ［315/410］ and 99.08% ［108/109］ vs 38.03% ［143/376］; all P＜0.01). The prevalence of NAFLD in T2DM patients with poor glycemic control was significantly higher than that with good glycemic control (69.33% ［495/714］ vs 36.95% ［75/203］, P＜0.001). The prevalence of NAFLD in T2DM patients with normal 25-hydroxyvitamin D levels was significantly higher than that with 25-hydroxyvitamin D inadequacy or deficiency (79.90% ［163/204］ vs 56.99% ［367/644］ and 57.97% ［40/69］, P＜0.001). After adjusting for confounding factors such as age, fasting blood glucose, blood glucose 2 h after meals, glycosylated hemoglobin (HbA1c), body mass index (BMI), systolic blood pressure, diastolic blood pressure, total cholesterol, triglyceride, high density lipoprotein-cholesterol, low density lipoprotein-cholesterol, aspartate aminotransferase (AST) and alanine aminotransferase (ALT), analysis of covariance showed that there were no significant differences in 25-hydroxyvitamin D levels between male and female NAFLD groups and non-NAFLD groups in either male or female patients (P＝0.065 and 0.197). The generalized linear model suggested that BMI, blood glucose 2 h after meals, HbA1c, homeostasis model of assessment for insulin resistance index, triglyceride and ALT/AST were the risk factors for NAFLD defined by ZJU index (OR＝2.754, 1.499, 1.341, 1.067, 2.871, and 22.346). There was no correlation between serum 25-hydroxyvitamin D level and ZJU index (P＝0.802). Conclusion ZJU index has application value in the diagnosis of NAFLD in T2DM patients. The prevalence of NAFLD defined by ZJU index is higher in T2DM patients with poor glycemic control, hypertension, hyperlipidemia, overweight or obesity, but serum 25-hydroxyvitamin D level is not associated with NAFLD.