Purpose The purpose of this paper is to analyse the clinical outcome data collected as part of an 18-week, abstinence-based residential therapeutic community (TC) programme, Higher Ground Drug Rehabilitation Trust (Higher Ground) in New Zealand. Lessons and implications for routine collection of clinical outcome data are identified. Design/methodology/approach Higher Ground collects longitudinal data on all consenting clients using a battery of validated psychometric tools, with repeated measures at up to nine points in time from first presentation through to 12-month post-discharge follow up. Data analysis covered clients who entered Higher Ground between 1 July 2012 and 2 June 2015 (n=524). Findings Clients presented with histories of addiction which often had significant negative associations with their physical and psychological health, their relationships, work, accommodation and criminal behaviour. By the time they exited the programme, clinically and statistically significant improvements were seen across multiple indicators including: substance use and abstinence; symptoms of post-traumatic stress disorder, depression, anxiety and stress; and a range of social indicators. Research limitations/implications Attrition in follow-up research is a significant challenge, with people completing the TC programme being more likely to participate than those who do not. This limits generalizability in post-discharge data. There was no control group, making causal attribution a challenge. Identifying suitable benchmarks from the literature is challenging because of the variety of outcome measures and research methodologies used. Practical implications Tracking client outcomes longitudinally using psychometric tools is potentially valuable for TCs and their funding bodies, as it provides insights into patterns of client recovery that can inform ongoing service improvements and resource allocation decisions. However, significant challenges remain. Originality/value The study demonstrates the value, and practical challenges, of collecting high-quality outcome data in a TC setting.