BackgroundThe crime rate in South Africa is extraordinarily high. The problem of crime is further complicated when a person, who suffers from a mental illness, becomes involved in a crime. Furthermore, the forensic evaluation of a person suspected of having a mental illness involved in alleged criminal behaviour can be challenging. However, a dearth of information exists in South African literature regarding the link between crime and mental illness.AimTo determine the percentage of alleged offenders, referred to the Free State Psychiatric Complex (FSPC) for observation, found accountable and not accountable, and to compare the biographical, diagnosis and offence profiles of these two groups. The analysis of differences can contribute to a better understanding of the complex process of forensic assessments.SettingForensic Observation Ward, FSPC, Bloemfontein.MethodsIn this comparative, retrospective study, all 505 trial-awaiting alleged offenders (observati) referred from 2009 to 2012 for a 30-day observation period, according to Sections 77 and/or 78 of the Criminal Procedures Act, were included. Results were summarised as frequencies and percentages, and means or percentiles. Significant differences between the groups were determined by sample t-tests or chi-squared tests.ResultsObservati found not accountable were in the majority (64.5%). Significant differences were found regarding marital and employment status, substance abuse, type of offence and diagnoses between the two groups. Almost all of the observati found to be not accountable were diagnosed with mental illness at the time of the assessment, whereas most observati found to be accountable for their actions at the time of the alleged offence were not found to be mentally ill. Observati found not accountable were significantly more likely to be diagnosed with schizophrenia, intellectual disability and substance-induced psychotic disorder, and committed mostly assault, murder and vandalism. Observati found accountable committed mostly rape, murder and theft.ConclusionThe majority of observati were found not accountable, with significant differences found between the two groups regarding demographic characteristics, type of diagnosis and offences committed. The identified differences can be used to assist in establishing criteria for the appropriate referral of alleged offenders by courts. Unnecessary referrals have a serious financial impact on the Department of Health. Furthermore, the high incidence of substance abuse among persons referred to the FSPC highlights the need for more substance rehabilitation centres in the Free State Province.