Abstract Background/Aims In rheumatoid arthritis (RA), musculoskeletal ultrasound (MSK-US) has greater sensitivity for the detection of active disease than clinical assessment using the DAS28 score, which is prone to false elevation by a spuriously high tender joint count (TJC) and/or visual analogue score (VAS) from non-RA causes. Thus, DAS28 may correlate poorly with MSK-US findings. In pregnancy, standardised disease activity of RA comprises calculation of the modified DAS28(3)CRP score, with removal of the VAS and ESR, both of which are confounded by pregnancy, and replacement with CRP; however, it is unknown whether non-specific musculoskeletal pain and/or peripheral oedema in (especially late) pregnancy might contribute to the TJC or swollen joint count (SJC) respectively and thus render this adapted score less reliable. Therefore, we conducted this study of pregnant RA patients and controls to assess the ability of the DAS28(3)CRP to detect disease activity compared with MSK-US. Methods Pregnant RA patients were recruited from the UCLH obstetric rheumatology clinic from September 2018 to March 2020. Patients underwent clinical assessment with DAS28(3)CRP and MSK-US longitudinally through pregnancy/post-partum. A control group of age-matched non-pregnant female RA patients was recruited from general rheumatology clinics and had equivalent assessments. MSK-US was undertaken using a Logiq S8. The standard protocol comprised 22-joint assessment of hands (dorsal longitudinal and transverse views of wrists, metacarpophalangeal and proximal interphalangeal joints). Quantification of Power Doppler (PD) signal was made as per the Outcome Measures in Rheumatoid Arthritis Clinical Trials (OMERACT) US definitions. PD scores were calculated as mean scores of all joints scanned. Statistical correlations were performed in SPSS using Spearman’s rank coefficient as data were non-normally distributed. Results A total of 45 MSK-US scans of pregnant/postpartum RA patients (N = 24, mean age 34.5+-3.6) were performed and compared to a control group of non-pregnant RA patients, who were scanned once each (N = 18, mean age 32.3+-5.2; p = 0.12). In the second (T2) and third trimester (T3) and the postpartum (PP), there was a significant positive correlation between DAS28(3)CRP and PD score (T2, ρ = 0.87 (p < 0.01); T3, ρ = 0.76 (p < 0.01); PP, ρ = 0.68 (p = 0.03)). Interestingly, each of these were stronger correlations than the equivalent comparison in non-pregnant RA patients (ρ = 0.51 (p = 0.03)). TJC also significantly positively correlated with PD score in each trimester of pregnancy and postpartum, but not in non-pregnant patients; difference between correlation coefficients was statistically significant for each trimester (T1, p < 0.01; T2, p = 0.05; T3, p = 0.05). Conclusion Interestingly, we found that that DAS28(3)CRP correlates better with PD signal on MSK-US in pregnancy than in non-pregnancy. Specifically, there was significantly better correlation of the TJC with PD score. It is unclear why non-inflammatory contributors to the TJC may be less prominent in pregnancy and further investigation is warranted. Disclosure C. Raine: Grants/research support; C.R. has received research funding from UCB. J. Manson: None. C. Ciurtin: None. I. Giles: Grants/research support; IG has received an unrestricted educational grant, speaker's fees and travel fees from UCB.