Performance usually improves when observers train with one type of a visual stimulus. Roving denotes the situation when, instead of one, two or more types of stimuli are presented randomly interleaved (one per trial). For some stimulus types, performance improves also in roving situations whereas for others it does not. To understand when roving impedes perceptual learning, we conducted four experiments. In each experiment, the very same bisection stimulus was randomly interleaved with a different stimulus type. Performance improved when a bisection and a vernier stimulus were randomly interleaved. However, no learning was found when we randomly interleaved this bisection stimulus with another one being twice as long. Based on these results, we propose that roving impedes performance when two stimuli excite strongly overlapping neural populations