Scientists debate how the public would react when anti-free will messages are communicated in the media. Taking part in this debate, social psychologists test whether letting participants read texts that argue against the plausibility of free will reduces their beliefs related to free will. However, this research leaves open whether merely reading such a text actually changes free will beliefs over and beyond a single experimental session in the lab. To fill this gap, in three experiments we investigated whether merely reading a short text claiming that free will does not exist, influences beliefs related to free will over time. Two experiments demonstrate that reading an anti-free will text (as compared to a control text) reduces participants' beliefs related to free will when measured on the same day (Experiment 1 and 2), the next day (Experiment 1 and 2) as well as one week later (Experiment 2). Critically, when controlling for potential demand effects by making sure that participants do not associate the assessment of free will beliefs with the first session in which beliefs were manipulated (Experiment 3), results no longer showed an effect of reading the anti-free will text over time. Theoretical and ethical implications of these findings are discussed.