With an increasing body of research on the roles of cognitive factors in SLA, the status of verbal reports as a cognitive measure is beginning to be established. However, little SLA research has assessed their validity in terms of reactivity (the effect of verbalisation on learners' performance). The current study investigates whether stimulated recall is reactive when it precedes post-tests and if so, what causes reactivity: recall stimuli, verbalisation or both. Forty-four learners of Japanese were randomly assigned to a stimulated recall, stimulus, experimental control or test control group. All groups participated in pre/post-testing. The first three groups also participated in communicative activities. Following the activities, the stimulated recall group participated in stimulated recall where they recalled their thoughts while viewing videotaped activities. The stimulus group also viewed a video of their activities, but did not recall their thoughts afterward. The experimental control group neither viewed a video nor recalled their thoughts. The results indicated non-significant differences between these three groups, suggesting the non-reactivity of stimulated recall, stimuli and verbalisation. However, it is suggested that the reactivity of stimulated recall cannot be discussed independently of various factors, such as the nature of stimuli, type of verbalisation and learner-interviewer interaction.