Three studies were conducted to assess the reliability and validity of a measure that we developed using a non-clinical sample of university undergraduates, the Driving and Riding Avoidance Scale (DRAS). Study 1 indicated that the scale was internally consistent (alpha=0.92) and that a four-factor model (general avoidance, avoidance of traffic and busy roads, avoidance of weather or darkness, and riding avoidance) provided the best fit to the data in a sample of 386 crash survivors. This study also revealed that survivors who received medical treatment for their crash-related injuries reported significantly greater avoidance than people who were uninjured or injured and not medically treated. Study 2 revealed that the DRAS possessed acceptable test-retest reliability (r=0.83) over a 4-week interval in a sample of 67 crash survivors. Using a sample of 118 survivors, study 3 examined the instrument's convergent and divergent validity through correlations with the Accident Fear Questionnaire (AFQ), the Mobility Inventory (MI), the Fear Survey Schedule-II (FSS-II), and the Fear Questionnaire (FQ). The strongest relationships were observed between the DRAS and the AFQ and with a driving subscale created from the MI items. The DRAS exhibited significantly weaker relationships with the FQ subscales that assessed other kinds of phobic avoidance.