This thesis analyses health services data for a group of 23,145 patients who were warned by their physicians regarding motor vehicle driving risks between April 1, 2006 and March 31, 2011. A three-year baseline interval prior to the warning was compared to the one year subsequent to the warning for the rate of involvement in motor vehicle crashes and for psychiatric hospitalization. The relative risk of involvement in a serious crash in the subsequent compared to the baseline interval was 0.69 (95% confidence interval 0.59 to 0.81, P<0.001). The relative risk of psychiatric hospitalization in the subsequent relative to the baseline interval was 1.97 (95% confidence interval 1.91 to 2.03, P<0.001). These results suggest that physician warnings can reduce subsequent involvement in serious crashes but are also associated with unintended adverse consequences. An awareness of these findings might inform clinical care and public health policy.