Background <p>To explore the transitions of different blood glucose states based on a multi-state Markov model among the Chinese health check-up population.</p> Methods <p>In this research, the multi-state Markov model was built based on 4776 blood glucose status assessments in 1677 Chinese participants. A quantitative longitudinal study was conducted to explore diabetes progression, providing information on the transition probability, the sojourn time, and the risk factors in three blood glucose states, namely normal (state 1), prediabetes (state 2), and diabetes (state 3).</p> Results <p>Among 4776 transition cases, there were 318 cases (6.7%) with remission (state 2→1), 3505 cases (73.4%) with stable disease (state 1→1, 2→2, 3→3), and 953 cases (19.9%) with progressive disease (state 1→2, 1→3, 2→3). The probability of progression from normal state to diabetes was 1.1% within 1 year and increased to 11.9% within 6 years. Increased age, blood pressure, and body mass index were risk factors for progression to diabetes in the normal state or prediabetes.</p> Conclusions <p>This article serves as a reference for targeted interventions for diabetes progression among the Chinese health check-up population. Obesity, hypertension and hyperlipidemia are risk factors that worsen the progression of diabetes, and preventive actions should be taken at an early stage.</p> Clinical trial number <p>Not applicable.</p>