Informing patients about treatment side effects increases the occurrence and intensity of side effects. Since the obligatory informed consent procedure in drug treatments requires transparency and nocebo research suggests that the informed consent of a drug leads to an increased occurrence of the mentioned side effects, the aim of this proof of concept study was to determine the effect of two different framings of informed consent on the occurrence, intensity, and perceived threat of side effects. Healthy male participants (n = 80) were randomized to one of two framing groups. The positive framing group was informed that the common side effect dizziness was a sign that the drug had started to work, while the neutral framing group was told that dizziness is an unpleasant but well-known side effect. Side effects were measured after the administration of metoprolol, an antihypertensive agent. Post hoc moderator analyses investigated the effect of pre-existing negative beliefs about the general harm of medication on the framing manipulation. Metoprolol-specific drug-attributed side effects were rated significantly less threatening in the positive framing group. The between-group effect size (Cohen's d) was small (d = 0.38, p = 0.049). Exploratory post hoc moderator analyses suggest that participants who believed that medication is a source of harmful effects benefited from positive framing, compared to neutral framing of drug-attributed side effects. Positive framing was partially effective in decreasing specific side effect measures, particularly among participants with a tendency to believe that medicine is harmful. Informed consent procedures should therefore be personalized, focusing on patients with negative treatment beliefs.