The present study explores the relationship between working memory capacity (WMC) and think-alouds, focusing on the issue of reactivity. Two WM span tasks (listening span and operation span) were administered to 42 English-speaking learners of Spanish. Learner performance on reading comprehension and written production was measured under two experimental conditions (think-aloud vs. non-think-aloud conditions). Results showed that think-alouds had negatively affected learner performance on reading comprehension, indicating the presence of reactive effects. Particularly interesting is the finding that reactive effects of think-alouds seem to have occurred in the course of rule learning among the high-WMC learners, but not among the low-WMC learners. The findings suggest that individual differences in WMC should be taken into careful consideration in future research that involves think-aloud protocols.