Background Schizophrenia is a debilitating mental health condition affecting the lives of many South Africans. The origins of the heterogeneity in the presentation of the illness remain uncertain. Aim This cross-sectional study performed a retrospective data analysis to determine the usefulness of digit ratio as an endophenotype in a South African schizophrenia population. Setting A large genetic study in a South African schizophrenia population recruited patients from services in the Western and Eastern Cape. Methods Complete clinical histories were captured for participants, including sets of images of the face and extremities. Software was utilised to measure the lengths of participants’ digits from said images and digit ratios (2D:4D) were calculated. Descriptive analyses were performed on the ratios and statistical differences in digit ratio means were calculated between groups characterised by sex, age of onset and the presence vs absence of positive symptoms. Linear modelling was utilised to assess for correlates between 2D:4D and positive and negative symptom severity using scores obtained from the Positive and Negative Syndrome Scale (PANSS) and Scale for the Assessment of Negative Symptoms (SANS). Results 2D:4D in male participants did not significantly differ from female participants as in healthy populations. 2D:4D did not significantly correlate with the severity of positive or negative symptoms and 2D:4D means between groups did not significantly relate to age of onset. Conclusion 2D:4D appears to be a possible endophenotype in schizophrenia in this population. 2D:4D, however, may not be as readily identifiable as certain minor physical anomalies and neurological soft signs significantly associated with schizophrenia in this population.