To assess pregnancy and fetal outcomes after in utero exposure to interferon-β (IFNβ) in all pregnancies occurring in women with multiple sclerosis (MS) during the study period, with a specific focus on the risk of spontaneous abortion. In this cohort study, data were gathered through a standardized, semi-structured interview. Patients who discontinued IFNβ less than 4 weeks from conception (exposed) were compared with those who had discontinued the drug at least 4 weeks from conception or who were never treated (not exposed). Possible confounders were handled through multivariate analyses adjusted for propensity score (PS). We collected data on 396 pregnancies in 388 women, 88 classified as exposed (mean exposure 4.6 ± 5.8 weeks). IFNβ exposure was not associated with an increased risk of spontaneous abortion (PS-adjusted odds ratio OR 1.08, 95% confidence interval CI 0.4 to 2.9, p = 0.88), although it was associated with both lower baby weight (PS-adjusted β -113.8, p < 0.0001) and length (PS-adjusted β -1.102, p < 0.0001). Proportion of spontaneous abortion in exposed patients fell within the range expected for the Italian population in the same period. IFNβ exposure (PS-adjusted OR 2.11, 95% CI 1.18 to 3.78, p = 0.012) and cesarean delivery were the only predictors of preterm delivery. In the exposed group, we did not observe any significant fetal complications, malformations, or developmental abnormalities over a median follow-up of 2.1 years. Our findings point to the relative safety of IFNβ exposure times of up to 4 weeks and can assist neurologists facing therapeutic decisions in women with MS with a pregnancy plan.