Background Persons with intellectual disabilities (ID) constitute a special‐needs population at high risk of falling. This is the first study to evaluate whether obstacle course training can improve mobility and prevent falls in this population. Methods The intervention was implemented as part of an institution‐wide health care improvement plan aimed at reducing falls at a residential facility for people with ID. It comprised an annual screening of each resident for his or her individual fall risk. Subsequently, the group of ambulatory persons with a moderate to high fall risk (n = 39) were offered 10‐session obstacle course training to improve their balance and gait abilities. Mobility was assessed pre‐intervention, mid‐term and post‐intervention with the Performance Oriented Mobility Assessment (POMA), the Timed Up and Go (TUG) and the 10‐meter walking test. The number of falls was compared between the year before and after intervention. Results The number of falls decreased by 82% (P < 0.001). POMA scores significantly improved from pre‐intervention to mid‐term (mean difference ± SD, 1.8 ± 2.9, P = 0.001), from mid‐term to post‐intervention (2.0 ± 2.9, P < 0.001), and from pre‐intervention to post‐intervention (3.8 ± 4.3, P < 0.001). Participants completed the 10‐meter walking test faster at the post‐intervention compared with the pre‐intervention assessment (difference ± SD, 2.1 ± 5.1 s, P = 0.022). TUG scores did not improve significantly. Conclusions The present study provides preliminary evidence for the effectiveness of obstacle course training in improving mobility and preventing falls in people with ID. As falls are a significant health concern in this population, further research is advocated to provide conclusive evidence for the suggested beneficial effects of exercise interventions.