AIM: To investigate the association between baseline serum 25-(OH)D3 level and the risk of metabolic syndrome in middle-aged and elderly people in Lanzhou. METHODS: A total of 2 754 non-MS people were selected from three communities in Lanzhou in 2011 to establish a cohort study population, In 2014, 2 309 subjects with complete data from two surveys were followed up for an average of 3.1 years. According to the quartile level of 25-(OH)D3, the baseline population was divided into Q1, Q2, Q3 and Q4 groups. The incidence of MS at different levels of 25-(OH)D3 was compared, and the correlation between 25-(OH)D3 and MS was analyzed. RESULTS: With the increase of baseline serum 25-(OH)D3 level, the prevalence of MS decreased (P<0.05). Baseline serum 25-(OH)D3 level was negatively correlated with WC, TG, FPG, 2h-PG, HbA1c and SBP (P<0.01).Logistic regression analysis showed that baseline serum 25-(OH)D3 level is an independent protective factor for the increase of WC, TG, FPG and MS (P<0.05). After an average follow-up of 3.1 years, a total of 341 patients developed MS (14.8%). Cox regression analysis showed that baseline serum 25-(OH)D3 level was correlated with the risk of abnormal HDL-C and 2 h-PG after follow-up (P<0.05). CONCLUSION: In the middle-aged and elderly population in Lanzhou, Gansu Province, baseline serum 25-(OH)D3 level is an independent protective factor for the increase of WC, TG, FPG and MS, the lower serum 25-(OH)D3 level is associated with the risk of decreased HDL-C level after follow-up. It has not been found that the baseline serum 25-(OH)D3 level is associated with the risk of MS and other components.