Individuals who do not fulfill required visual field standards have their driving license withdrawn in Sweden. However, understanding of the ability to compensate for this loss is limited. This study aimed to determine if reaction time and cognitive performance are important for safe driving in visual field loss (VFL) individuals. Visually demanding reaction time tasks of different complexity, for example, can help one understand why some VFL individuals drive as safely as normally sighted individuals. Twenty VFL individuals and 83 normally sighted individuals participated in a driving simulator experiment and an additional test battery. The driving task categorized VFL participants into two subgroups: passed or failed. Three reaction time tasks, four cognitive tests, and two meta-cognitive scales were completed. The passed VFL subgroup was faster than the failed subgroup in the context-dependent reaction time task and slower in the context-independent reaction time task. The passed subgroup performed equally well, or less well, on the cognitive tasks compared to the failed subgroup. The VFL participants performed less well than the normally sighted individuals on most cognitive tasks. However, VFL participants did not reflect on their driving ability (in meta-cognitive scales) in the same way as normally sighted individuals. There appear to be VFL subgroups in terms of ability to drive safely. Reaction time is important, but context dependent. Cognitive context-independent tests appear unrelated to driving test outcome for VFL individuals. The problems with context-independent testing of perceptual, cognitive, and meta-cognitive abilities when predicting safe driving capabilities are discussed. © 2019, The Author(s).