Although negative effects of psychological interventions are suspected to be common, they are rarely investigated. Experts and international guidelines agree that monitoring for negative effects in clinical studies is needed to make psychological interventions safer and to empower patients before they give their consent to participate. Therefore, monitoring should already be considered during planning and preregistration of a study. The aim of this scoping review was to find out how frequently studies on psychological interventions monitor their negative effects according to preregistrations and to investigate reasons why monitoring is not carried out. Preregistrations of psychological interventions on ClinicalTrials.gov were scrutinized for information on monitoring of negative effects and other study characteristics. In a survey, researchers of studies where no monitoring was reported were asked for reasons for not doing so. Overall, 2231 preregistrations of psychological interventions were found; of these, only 3.4% included explicit information on monitoring for negative effects. In the survey, more researchers reported having conducted monitoring, although the type of monitoring was often inadequate. The type of monitoring varied widely, and specific monitoring measurements were rarely used repeatedly. Monitoring for negative effects was more prevalent in studies investigating treatments versus low‐threshold interventions, in studies conducted in Europe versus other continents and in more recent studies. Researchers reported lack of knowledge as the most frequent reason for not monitoring negative effects. Our results imply a lack of monitoring and inconsistent information on negative effects in preregistrations, with inconsistent use of the term monitoring and measurements, and a lack of knowledge among researchers. Improved knowledge and a standardized approach, starting with an adequate preregistration, would be helpful to routinely examine negative effects in psychological interventions to make them safer and better.