Background: Given that fewer than 50% of countries provide Opioid Agonist Maintenance Therapies (OAMT), it is important to assess whether other substances act as a substitute for heroin in recovering heroin users who receive detoxification models of treatment. There is a dearth of prospective studies from low-and-middle-income countries evaluating these patterns of substance use. Methods: 300 heroin users from the Gauteng province of South Africa were assessed on entry into inpatient detoxification and then followed-up 3 and 9 months after leaving treatment. Treatment consisted of 1 week of detoxification followed by 6-8 weeks of psychosocial therapy. We measured the overall changes in the prevalence of heroin, alcohol and other drug use at baseline and postrehabilitation. Comparison of these outcomes at enrolment, 3 months and 9 months was performed by a Generalised Estimating Equation (GEE) with the outcome as the dependent variable, observation point as the independent variable, and participant as the repeated measure. Injecting status and treatment completion were included as covariates. We also measured the individual pathways between heroin and alcohol use in the 210 participants that were seen at all three timepoints. Results: Of the original cohort, 252 (84.0%) were re-interviewed at 3 months and 225 (75.0%) at 9 months. From baseline to 3 months, the proportion of past month heroin users decreased significantly to 65.5%; however, during this time, the proportion of past month alcohol users increased from 16.3% to 55.2% (p<0.0001). When assessing the pathways between heroin and alcohol use at an individual level, 55.4% (n-97) of those who were past month alcohol abstinent prior to rehabilitation were using alcohol at 3 months. From 3 to 9 months the proportion of heroin users increased to 72.4% (p<0.0001), and during this time, the proportion of alcohol users decreased. Conclusion: After detoxification, a significant reduction in heroin use was observed with a concomitant increase in alcohol consumption. Under these circumstances, alcohol may have acted as a substitute for heroin in the short term. The initial reduction in heroin use 3 months postrehabilitation was followed by increased consumption 6 months later. This observation supports the need for interventions to prevent, monitor and treat high levels of alcohol use in heroin users post detoxification. The provision of OAMT is a necessary consideration to address both the risk of increased alcohol intake as well as the decline in heroin abstinence rates.