Abstract: Objective To investigate the relationship between arterial stiffness and new-onset diabetes mellitus in hypertensive population. Methods The subjects of this study were from the Kailuan Study population. Arterial stiffness was evaluated by brachial-ankle pulse wave velocity（baPWV）. A total of 8 061 hypertensive and non-diabetic employees who participated in the 2010-2011, 2012-2013 and 2014-2015 year physical examination of the Kailuan Study were eligible for this study. Baseline was set at the first baPWV measurement, while follow-up ended at the first incident of diabetes mellitus or until the last examination（2016-2017 year）. According to baPWV quartile,the effect of baPWV on new-onset diabetes mellitus was analyzed by Cox proportional hazard model, a sensitivity test was analyzed by body mass index（BMI） stratification. Results The mean follow-up time was（4.22±2.10） years. During the follow-up period, 874（10.8%） participants developed diabetes, with an average incidence rate of 25.71 per 1 000 person-years. The incidence rate（per 1 000 person-years） of new-onset diabetes in the first to fourth quartile of baPWV were 14.60, 23.77, 27.15 and 37.70, respectively（P＜0.001）. Compared with the first quartile of baPWV, after adjustment for age, gender, mean arterial pressure, baseline fasting glucose, hypersensitive C-reactive protein and other confounding factors, the HR and 95% CI for new-onset diabetes of the second, third and fourth quartile were 1.57（1.24-1.98）, 1.74（1.37-2.20） and 2.38（1.87-3.03）, respectively. For each standard deviation increase of baPWV, the HR and 95% CI was 1.32（1.24-1.42）. Conclusion Arterial stiffness is associated with new-onset diabetes mellitus in hypertensive population.