Background: Several studies have shown that pregnancy reduces multiple sclerosis (MS) relapses, which increase in the early postpartum period. Postpartum relapse risk has been predicted by pre-pregnancy disease activity in some studies. Objective: To re-examine effect of pregnancy on relapses using the large international MSBase Registry, examining predictors of early postpartum relapse. Methods: An observational case-control study was performed including pregnancies post-MS onset. Annualised relapse rate (ARR) and median Expanded Disability Status Scale (EDSS) scores were compared for the 24 months pre-conception, pregnancy and 24 months postpartum periods. Clustered logistic regression was used to investigate predictors of early postpartum relapses. Results: The study included 893 pregnancies in 674 females with MS. ARR (standard error) pre-pregnancy was 0.32 (0.02), which fell to 0.13 (0.03) in the third trimester and rose to 0.61 (0.06) in the first three months postpartum. Median EDSS remained unchanged. Pre-conception ARR and disease-modifying treatment (DMT) predicted early postpartum relapse in a multivariable model. Conclusion: Results confirm a favourable effect on relapses as pregnancy proceeds, and an early postpartum peak. Preconception DMT exposure and low ARR were independently protective against postpartum relapse. This novel finding could provide clinicians with a strategy to minimise postpartum relapse risk in women with MS planning pregnancy.