Statistical learning (SL) is how humans identify patterns in the world. Two key aspects of SL are how much a person can learn with practice (potential) and how quickly they improve (efficiency). We used mathematical models to measure these two aspects separately and studied how different mental abilities influence them. Our findings showed that people with high potential for SL tend to have poor inhibition (the process of stopping impulses) but good visuo-spatial working memory (the ability to remember and manipulate visual information). In contrast, high efficiency in SL is linked to good inhibition and good set-shifting (ability to switch attention between tasks). We discovered that potential and efficiency in SL are influenced by different cognitive functions.