Summary:As the population has become both older and more technologically literate, a new class of "brain training" computer programs have gained in popularity. Though these programs have attracted substantial attention from scientists and consumers, the extent of their benefits, if any, remain unclear. Here we employ neuropsychological tests and behavioral metrics collected during periods of real-world driving (with and without manipulations of cognitive load) to evaluate the effects of training with Posit Science's DriveSharp software. We find that DriveSharp's training effects appear in in-lab measures of Useful Field of View but did not translate to changes in actual driving performance or changes in visual behavior in consistent or quantifiable ways in the sample assessed. The implications of these results and relevant limitations of the present research are discussed.