Previous research has shown that children with autism spectrum disorder (ASD) can categorize visual stimuli without direct training when they can also tact these stimuli using a common name and behave as listeners in relation to this name. However, children usually learn to assign objects specific names prior to learning the category to which they belong. The current study replicated previous research and evaluated whether multiple‐tact training would establish visual categorization (measured by a picture sorting test) and listener behavior. We used a nonconcurrent multiple baseline design across 2 children with autism spectrum disorder. After multiple‐tact training, we assessed whether participants would visually categorize stimuli based on their common category name. Both participants categorized and engaged in the corresponding listener behavior.