Background Serious mental illness (SMI), including schizophrenia, bipolar disorder and other psychoses, is linked with high disease burden, poor outcomes, high treatment costs and lower life expectancy. In the UK, most people with SMI are treated in primary care by general practitioners (GPs), who are financially incentivised to meet quality targets for patients with chronic conditions, including SMI, under the Quality and Outcomes Framework (QOF). The QOF, however, omits important aspects of quality. Objective(s) We examined whether better quality of primary care for people with SMI improved a range of outcomes. Design and setting We used administrative data from English primary care practices that contribute to the Clinical Practice Research Datalink GOLD database, linked to Hospital Episode Statistics, Accident & Emergency (A&E) attendances, Office for National Statistics mortality data, and community mental health records in the Mental Health Minimum Dataset. We used survival analysis to estimate whether selected quality indicators affect the time until patients experience an outcome. Participants Four cohorts of people with SMI depending on the outcomes examined and inclusion criteria. Interventions Quality of care was measured with: i) QOF indicators: care plans and annual physical reviews ;and ii) non-QOF indicators identified through a systematic review (antipsychotic polypharmacy and continuity of care provided by GPs). Main outcome measures Several outcomes were examined: emergency admissions for i) SMI and ii) ambulatory care sensitive conditions (ACSCs); iii) all unplanned admissions; iv) A&E attendances; v) mortality; vi) re-entry into specialist mental health services; vii) costs attributed to primary, secondary and community mental healthcare. Results Care plans were associated with lower risk of A&E attendance (Hazard ratio (HR) 0.74, 95%CI 0.69-0.80), SMI admission (HR 0.67, 95%CI 0.59-0.75), ACSC admission (HR 0.73, 95%CI 0.64-0.83), and lower overall healthcare (£53), primary care (£9), hospital (£26), and mental healthcare costs (£12). Annual reviews were associated with reduced risk of A&E attendance (HR 0.80, 95%CI 0.76-0.85), SMI admission (HR 0.75, 95%CI 0.67-0.84), ACSC admission (HR 0.76, 95%CI 0.67-0.87), and lower overall healthcare (£34), primary care (£9), and mental healthcare costs (£30). Higher GP continuity was associated with lower risk of A&E presentation (HR 0.89, 95%CI 0.83-0.97), ACSC admission (HR 0.77, 95%CI 0.65-0.92), but not SMI admission. High continuity was associated with lower primary care costs (£3). Antipsychotic polypharmacy was not statistically significantly associated with the risk of unplanned admission, death or A&E presentation. None of the quality measures were statistically significantly associated with risk of re-entry into specialist mental healthcare. Limitations There is risk of bias from unobserved factors. To mitigate this, we controlled for observed patient characteristics at baseline and adjusted for the influence of time-invariant unobserved patient differences. Conclusions Better performance on QOF measures and continuity of care are associated with better outcomes and lower resource utilisation and could generate moderate cost savings. Future work Future research should examine the impact of primary care quality on measures that capture broader aspects of health and functioning.