It is known that both cognitive training and aerobic exercise training can independently improve cognitive performance in older adults. Combined multimodality (cognitive and aerobic) training has shown promise in producing significantly more training gains in older adults than pure physical exercise and control; however, results are mixed. To address these mixed results and elucidate the efficacy of different schedules of multimodal training, we compared simultaneous and sequential formats. To this end, 42 older adults (M = 68.05 years) participated in 12 sessions of same-day multimodal training and were randomly assigned to either the simultaneous (concurrent cognitive dual-task and aerobic exercise) or sequential training group (cognitive dual-task followed by aerobic exercise). Both groups showed significant improvement on measures of processing speed and verbal memory following training, with the sequential group showing a significant training advantage on working memory. Motivation to engage in cognitive effort moderated the training gains in verbal memory, and baseline aerobic fitness moderated the magnitude of training gains in response inhibition negatively. The current findings demonstrate the selective effects of training format on cognitive outcomes, in line with the principle of neural overlap, which suggests that improvements should occur to the extent that the trained and untrained tasks share underlying neural pathways. The results also underscore the importance of considering individual difference factors in cognitive training studies.