BACKGROUND: We aimed to explore the occurrence and persistence of symptoms, diagnoses and prescribing after COVID-19 among populations from earlier (wave 2) and later (wave 4) in the pandemic. METHODS: With the approval of NHS England, we analysed data from English primary care using The Phoenix Partnership SystmOne through the OpenSAFELY data analytics platform. Individuals with community-diagnosed COVID-19 September 2020-January 2021 (wave 2) were matched to contemporary (2020-2021) and historical (2017-2018) comparators. Individuals with community COVID-19 December 2021-March 2022 (wave 4) were matched to contemporary comparators (last follow-up 31 March 2023). Occurrence of each of (1) long-COVID symptoms; (2) primary-care diagnoses and (3) new prescriptions was analysed at any time during 1 year after COVID-19 and at: 4-12 weeks, 12 weeks-6 months and 6 months-12 months after COVID-19 to assess persistence. RESULTS: 902 885 COVID-19 cases (wave 2) matched to 4 449 265 contemporary (no-COVID-19) comparators. 1 553 160 COVID-19 cases (wave 4) matched to 7 624 770 contemporary comparators. Positive wave 2 associations after COVID-19 were observed for hair loss (OR 1.57, 95% CI 1.48 to 1.66), mobility impairment (1.41, 1.35 to 1.48), fatigue (1.46, 1.42 to 1.49), cognitive impairment (1.39, 1.34 to 1.44) and loss of taste or smell (1.38, 1.31 to 1.46). At 6-12 months reporting persisted for mobility impairment, fatigue and cognitive impairment. There were small increases in new prescriptions for NSAIDs (1.24, 1.23 to 1.26), drugs to treat infections (1.24, 1.23 to 1.25) and musculoskeletal problems (1.23, 1.22 to 1.25). Wave 4 associations were generally weaker than Wave 2. CONCLUSIONS: Long-COVID symptoms and new prescribing generally reduce over time and are potentially less problematic following less severe illness. Fatigue/cognitive/mobility symptoms persist following COVID-19.