Abstract Background/Aims To identify the risk of serious infections (SI) according to initial treatment strategy, using conventional synthetic disease modifying anti-rheumatic drugs (csDMARD) and corticosteroids, in patients recruited to the National Early Inflammatory Arthritis Audit (NEIAA). Methods An observational cohort study design was used. The population included adults in England with a new onset rheumatoid arthritis (RA), fulfilling ACR/EULAR 2010 criteria, between April 2018-March 2021. Outcomes studied were SI, defined by infections requiring hospitalisation (primary admission diagnosis/nosocomial acquisition) or death (SI stated on death certificate), identified using NHS Digital linkage. Patients' characteristics were tabulated by treatment strategies. Hazard ratios (HR) were calculated using single failure Cox proportional-hazards models, with confounders-adjusted models (age, gender, smoking status, comorbidities, social deprivation) and fully-adjusted models including disease factors (seropositivity, DAS28). Individuals were considered at risk from the date of RA diagnosis, and censored at SI event, death, or March 2021 (whichever was earliest). Results 20,060 patients with RA were included. Initial DMARD therapy was known for 19,572 patients, of whom 11,966 were on methotrexate/MTX based strategies (mono or combo), 5,059 on csDMARD combination strategies (other than MTX) and 2,547 on no DMARD strategy. 15,319 patients were on corticosteroids at baseline. Mean age 59.5 years (+/-15); 63% female; smoking status (20% current; 30% ex-smokers); comorbidities (21% hypertension; 10% diabetes; and 12% lung disease). Rheumatoid Factor/CCP antibodies were positive in 68%. At presentation, median disease scores were 5.1 (interquartile range [IQR]: 4.0-5.9) for DAS28, 1.1 (IQR: 0.6-1.7) for health assessment questionnaire (HAQ) and 24 (IQR: 16.0-33.0) for musculoskeletal health questionnaire (MSKHQ).There were 519 SI admissions and 17 SI deaths, corresponding to incidence rates per 100 person-years for admissions: 3.19 (95% CI: 2.93-3.48) and deaths: 0.10 (95% CI: 0.06-0.16). In fully-adjusted models, increasing age predicted both SI admissions and deaths. Being a smoker, having a comorbidity, higher disease activity (DAS28), symptom burden (MSKHQ) and disability (HAQ) at presentation associated with more SI admissions. For each 1 unit increase in DAS28, the risk of SI increased by 8% (HR 1.08 [95% CI:1.01-1.16]). Seropositivity did not associate with SI. MTX-based strategies 0.75 (95% CI:0.62-0.91) and csDMARD combination therapy 0.70 (95% CI:0.53-0.94) associated with fewer SI admissions compared to no DMARD. In unadjusted models, corticosteroid associated with more SI admissions 1.29 (95% CI:1.10 -1.62); however, in fully-adjusted models this association was no longer statistically significant. csDMARD strategies did not associated with SI deaths in any of the models. Conclusion Patient and disease factors at diagnosis appear to be important predictors of admissions and mortality for serious infections. Infection risk appears to be greatest in those with higher RA disease activity. An important limitation is that NEIAA does not capture data on treatment changes over time and steroid use beyond baseline. Disclosure M. Adas: None. K. Bechman: None. M. Russell: Honoraria; Lilly and Menarini. Other; upport for attending conferences from Lilly, Pfizer, Janssen, and UCB, and advisory board fees from Biogen.. I. Karafotias: None. D. Nagra: None. S. Patel: None. S. Gallagher: None. E. Price: None. M. Garton: None. A. Rutherford: None. A. Cope: None. S. Norton: None. J. Galloway: Honoraria; from Abbvie, Biovitrum, Bristol Myers Squib (BMS), Celgene, Chugai, Gilead, Janssen, Lilly, Novartis, Pfizer, Roche, Sanofi, Sobi, and UCB.