BackgroundAddressing the wider determinants of mental health alongside psychological therapy could improve mental health service outcomes and population mental health.ObjectivesTo estimate the effectiveness of an enhanced ‘Improving Access to Psychological Therapies’ (IAPT) mental health service compared with traditional IAPT in England. Alongside traditional therapy treatment, the enhanced service included well-being support and community service links.DesignA real-world evaluation using IAPT’s electronic health records.SettingThree National Health Service IAPT services in England.ParticipantsData from 17 642 service users classified as having a case of depression and/or anxiety at baseline.InterventionWe compared the enhanced IAPT service (intervention) to an IAPT service in a different region providing traditional treatment only (geographical control), and the IAPT service with traditional treatment before additional support was introduced (historical control).Primary outcome measuresPatient Health Questionnaire-9 (PHQ-9) Depression Scale (score range: 0–27) and Generalised Anxiety Disorder-7 (GAD-7) Anxiety Scale (score range: 0–21); for both, lower scores indicate better mental health. Propensity scores were used to estimate inverse probability of treatment weights, subsequently used in mixed effects regression models.ResultsSmall improvements (mean, 95% CI) were observed for PHQ-9 (depression) (−0.21 to –0.32 to −0.09) and GAD-7 (anxiety) (−0.23 to –0.34 to −0.13) scores in the intervention group compared with the historical control. There was little evidence of statistically significant differences between intervention control and geographical control.ConclusionsEmbedding additional health and well-being (H&W) support into standard IAPT services may lead to improved mental health outcomes. However, the lack of improved outcomes compared with the geographical control may instead reflect a more general improvement to the intervention IAPT service. It is not clear from our findings whether an IAPT service with additional H&W support is clinically superior to traditional IAPT models.