Analogue samples are often used to study obsessive–compulsive (OC) symptoms and related phenomena. This approach is based on the hypothesis that results derived from such samples are relevant to understanding OC symptoms in individuals with a diagnosis of obsessive–compulsive disorder (OCD). Two decades ago, Gibbs (1996) reviewed the available literature and found initial support for this hypothesis. Since then there have been many important advances addressing this issue. The purpose of the present review was to synthesize various lines of research examining the assumptions of using analogue samples to draw inferences about people with OCD. We reviewed research on the prevalence of OC symptoms in non-clinical populations, the dimensional (vs. categorical) nature of these symptoms, phenomenology, etiology, and studies on developmental and maintenance factors in clinical and analogue samples. We also considered the relevance of analogue samples in OCD treatment research. The available evidence suggests research with analogue samples is highly relevant for understanding OC symptoms. Guidelines for the appropriate use of analogue designs and samples are suggested. •We reviewed evidence for utility of analogue samples in understanding OC symptoms.•OC symptoms are sufficiently prevalent in non-clinical samples.•OC symptoms are dimensional rather than categorical in frequency and severity.•Qualitative aspects of OC symptoms are similar in clinical and nonclinical samples.•Similar causal and maintenance factors occur in clinical and nonclinical samples.