PURPOSE: The purpose of this study was to investigate the relationship between visual field loss and driving performance as determined by on-road driving assessments. METHODS: We reviewed the files of 1350 patients enrolled in a rehabilitation program at the Bloorview MacMillan Rehabilitation Centre, Toronto, Canada. We identified 131 patients with visual field loss who had undergone an on-road driving assessment. These patients had a primary diagnosis of visual impairment or a primary diagnosis of cerebral vascular accident (CVA) with a secondary diagnosis of visual impairment. None of these patients had documentation of neglect, substantial motor or cognitive deficits. We report the data obtained from 13 hemianopics, 7 quadrantanopics, 25 patients with monocular vision, 10 patients with moderate peripheral losses (<135 degrees of horizontal visual field measured at the midline), and 76 patients with mild peripheral losses (between 135 degrees and 186 degrees of horizontal visual field). The on-road assessment consisted of driving in the area surrounding the rehabilitation center, and the outcome was based on performance on a number of tasks commonly encountered in daily driving. For the purposes of this study, the assessment outcomes were classified as safe, unknown, or unsafe. RESULTS: Overall, the extent of visual field loss did not have a significant impact on driving performance (chi2 = 4.37, p = 0.358). However, hemianopia tended to have a worse impact on driving performance than quadrantanopia with a marginally significant result (chi2 = 3.33, p = 0.068). Overall, the location of the visual loss was not significantly related to driving fitness (chi2 = 1.05, p = 0.30). However, localized defects in the left hemifield (chi2 = 9.561, p = 0.002) and diffuse visual loss in the right hemifield (chi2 = 10.395, p = 0.001) seemed to be associated with driving impairments. A large proportion of monocular drivers were safe drivers and the location of their deficit had no significant impact. CONCLUSIONS: Although the extent of visual field defects appears to be related to driving performance as determined by an on-road driving assessment, large individual differences were observed. This highlights the need for individualized on-road assessments for patients with visual field defects.