BACKGROUNDS: It has been hypothesized that ankylosing spondylitis is associated with an increased risk of incident hip fractures due to osteoporosis and risk of falls but the supporting evidence is limited and mixed. OBJECTIVES: To assess the risk of hip fractures in a large cohort of patients with ankylosing spondylitis compared to a matched cohort. DESIGN: A retrospective cohort study. SUBJECTS: Men and women diagnosed with ankylosing spondylitis from 1 January 2002 to 31 December 2018. Matching in a 5:1 ratio was based on age and sex. Followup ended on 23 June 2019. MAIN MEASURES: Cox regression models adjusting for confounders defined in a causal inference framework were used to determine the hazard ratio for hip fractures. KEY RESULT: The final cohorts included 5,909 ankylosing spondylitis patients and 28,671 matched patients. The ankylosing spondylitis cohort had a mean age of 49 (17) years and was composed of 3,762 (64%) men, 3,638 (62%) patients born in Israel, and 1,532 (26%) patients of low residential socioeconomic status. During 45,388 and 224,192 cumulative person-years of follow-up, the ankylosing spondylitis and matched cohorts had 2.47 and 1.63 cases of hip fractures per 1,000 person-years, respectively. Ankylosing spondylitis patients also developed hip fractures earlier (74 [13] vs. 79 [10] years, p = 0.002). Ankylosing spondylitis was associated with hip fractures in the unadjusted (HR = 1.52, 95% CI [1.23-1.88]) and adjusted (HR = 1.56, 95% CI [1.27-1.93]) models. The association was evident in men (HR = 1.65, 95% ) and women (HR = 1.48, 95% CI [1.07-2.05]). CONCLUSION: This study found that ankylosing spondylitis patients developed hip fractures earlier and more often compared to a matched cohort. This study suggests that ankylosing spondylitis patients might benefit from more proactive screening, mitigation, and prevention of risk factors for hip fractures.