Laboratory tasks such as the Reflexive Imagery Task (RIT) have revealed that external stimuli (e.g., line drawings) can elicit involuntary mental imagery (Allen, Wilkins, Gazzaley, & Morsella, 2013). In the present variant of the RIT, subjects (n = 40) were instructed to think of a random word when presented with a cue (a square frame). Subjects were instructed to disregard any image (e.g., a line drawing of a cat) that was presented inside the frame. Unlike in a standard RIT, subjects were instructed and encouraged to generate their own mental imagery. On half of the 120 trials, an image (selected from a set of 60 line drawings of everyday objects) was presented.On the other half of the trials, no image was presented. At the beginning of each trial, and after a pause of 700 ms, a bell sounded. After the sounding of the bell, the frame appeared, with or without an image, at one of three times randomly (1 second, 2 seconds, or 4 seconds). Across all conditions, subjects reported thinking of a word on the majority of the trials (the mean proportions across all conditions > .80). A fully within-subjects ANOVA with the factor Presence (Image Present or Image Absent) and the factor Pre-Stimulus Preparation Time (1 second, 2 seconds, or 4 seconds) revealed a main effect Presence, F (1, 39) = 10.48, p = .002 (ηp 2. = .212), and Pre-Stimulus Preparation Time, F (1.86, 72.41) = 9.64, p < .001 (ηp 2. = .198). There was also a significant interaction between the two factors, F (2, 77.81) = 6.54, p = .002 (ηp 2. = .144). The present finding illuminates under-investigated aspects of the RIT effect and can also shed light on the basic mechanisms involved in the generation of involuntary thoughts, thoughts that are often triggered by external stimuli and that can be debilitating in clinical disorders (e.g., rumination, addictions, and obsessions).