This article describes two power mobility training protocols used with seniors and compares posttraining driving performance. Twelve users of power mobility were consecutively recruited from two residential facilities in Toronto, Canada. The aim of training at both sites was to make clients comfortable with and safe at driving power mobility devices. The content of training was similar, but training protocols differed significantly in terms of the number of sessions (means of 3.43 vs. 9.80; p Յ .05) and the time frame over which the sessions were offered (means of 1.57 vs. 5.10 weeks; p Յ .01). Participants at the two sites differed significantly in terms of overall driving performance (p Յ .05), gender (p Յ .01), and type of device used (p Յ .05). Overall, driving performance was significantly associated with facility, gender, type of device used, and training duration (p Յ .05). When these variables were entered into an exploratory hierarchical regression, facility accounted for 64% of the variance in driving performance. When facility was controlled for, the correlations between device and duration of training with driving performance were no longer significant. The determinants of driving performance are difficult to clearly specify as the variable facility encompasses gender as well as all other differences between the two training protocols. Nevertheless, these data provide direction for future research in this area.