Aims: To present final efficacy/safety results from a prospective, long-term extension trial of onabotulinumtoxinA for urinary incontinence (UI) due to neurogenic detrusor overactivity (NDO); patients received treatment for up to 4 years. Methods: Patients who completed a 52-week, phase III trial of onabotulinumtoxinA for NDO were eligible to enter a 3-year, multicenter, open-label extension study of intradetrusor onabotulinumtoxinA (200U or 300U). Patients were treated ''as needed'' based on their request and fulfillment of prespecified qualification criteria (!12 weeks since previous treatment and a UI episode threshold). Assessments included change from study baseline in UI episodes/day (primary efficacy measure), volume/void, and Incontinence Quality of Life (I-QOL) total score (week 6); duration of effect; adverse events (AEs); and initiation of de novo clean intermittent catheterization (CIC). Data are presented for up to six treatments. Results: OnabotulinumtoxinA 200U consistently reduced UI episodes/day; reductions from baseline ranged from -3.2 to -4.1 across six treatments. Volume/void consistently increased, nearly doubling after treatment. I-QOL improvements were consistently greater than twice the minimally important difference (þ11 points). Overall median duration of effect was 9.0 months (200U). Results were similar for onabotulinumtoxinA 300U. Most common AEs were urinary tract infections and urinary retention. De novo CIC rates were 29.5, 3.4, and 6.0% (200U), and 43.0, 15.0, and 4.8% (300U) for treatments 1-3, respectively; de novo CIC rates were 0% for treatments 4-6. Conclusions: OnabotulinumtoxinA treatments consistently improve UI, volume/void, and QOL in patients with UI due to NDO in this 4-year study, with no new safety signals.