Abstract Objectives To explore the association of maternal RA to pregnancy outcomes, especially preterm birth (PTB) and small for gestational age (SGA), in relation to disease activity and anti-rheumatic treatment before and during pregnancy. Methods By linking prospective clinical rheumatology registers (CRR) in Sweden (the Swedish Rheumatology Quality Register, SRQ) and Denmark (the Danish clinical quality register, DANBIO) with medical birth registers, we identified 1739 RA-pregnancies and 17 390 control-pregnancies (matched 1:10 on maternal age, birth year, parity) with delivery 2006–18. Disease activity (DAS28, CRP, HAQ score) and anti-rheumatic treatment 9 months before and during pregnancy were identified through CRR and prescribed drug registers. Using logistic regression, we estimated adjusted odds ratios (aOR) with 95% CI for PTB and SGA overall and stratified by disease activity and anti-rheumatic treatment before and during pregnancy, adjusting for maternal characteristics. Results We found increased aOR of PTB [1.92 (1.56–2.35)] and SGA [1.93 (1.45–2.57)] in RA-pregnancies vs control-pregnancies. For RA-pregnancies with DAS28-CRP ≥4.1 vs <3.2 during pregnancy, aOR was 3.38 (1.52–7.55) for PTB and 3.90 (1.46–10.4) for SGA. Use of oral CS (yes/no) during pregnancy resulted in an aOR of 2.11 (0.94–4.74) for PTB. The corresponding figure for biologics was 1.38 (0.66–2.89). Combination therapy, including biologics before pregnancy, was a marker of increased risk of both PTB and SGA. Conclusion During pregnancy, disease activity rather than treatment seems to be the most important risk factor for PTB and SGA in RA. Women with RA should be carefully monitored during pregnancy, especially if they have moderate to high disease activity or/and are treated with extensive anti-rheumatic treatment.