Background. Anhedonia, defined as a reduced capacity to experience pleasure, has been associated with many clinical conditions, including major depressive disorder (MDD), schizophrenia (SCZ), substance use disorder (SUD) and Parkinson’s disease (PD). Anhedonia symptoms are rarely compared across conditions however, and it is currently unclear whether symptom severity differs between clinical groups. Reference values for hedonic capacity in healthy humans are also missing from the literature. Objectives. To generate and compare reference values for anhedonia levels across healthy and clinical groups, we performed a set of meta-analyses of self-reported anhedonia as measured with a widely used questionnaire, the Snaith-Hamilton Pleasure Scale (SHAPS). We also calculated prediction intervals for each group, providing the range of mean SHAPS scores to be expected in future studies. Methods. We extracted SHAPS scores from all available studies citing the initial scale development paper (189 papers) and used random-effects models to calculate average SHAPS scores and 95% confidence intervals separately for samples of healthy participants and samples of patients with current MDD, past/remitted MDD, SCZ, SUD and PD. We used meta-regression to compare SHAPS scores between these groups. Results. In the available literature, patients with current MDD, SCZ, SUD and PD all scored higher on the SHAPS than healthy participants. SHAPS scores in SCZ, SUD and PD were nevertheless considerably lower than scores in current MDD. Conclusion. Our results indicate that the severity of anhedonia differs across disorders that have been associated with anhedonia. Whereas anhedonia in current MDD likely affects multiple domains of pleasure (e.g. food/drink, pastimes/hobbies, social, physical), anhedonia in SCZ, SUD and PD may instead reflect a decrease in projected enjoyment of only a minority of life’s many rewards.