INTRODUCTION:Many previous studies have been limited by self- or proxy-reported injury or short follow-up. We investigated whether head or brain injuries are associated with Alzheimer's disease (AD), possible modifying factors and dose-response relationship. METHODS:Nested register-based case-control study of all community dwellers who received clinically verified AD diagnosis in Finland in 2005 to 2011 (n = 70,719) and one to four matched controls for each case (n of controls = 282,862). RESULTS:The magnitude of association between hospital-treated head and/or brain injuries was strongly dependent on the lag time between exposure and outcome. With a 5-year lag time, head injury (adjusted odds ratio; 95% confidence interval 1.19; 1.15-1.23) or brain injury (1.23; 1.18-1.29) was associated with higher risk of AD. Dose-response relationship with number and severity of injuries was observed. Associations were stronger in those with earlier onset of AD. CONCLUSIONS:Stronger associations with shorter lag times indicate that head and/or brain injuries may also reflect the ongoing AD disease process.