Fetuses with congenital heart disease (CHD) are at increased risk of pregnancy loss compared with the general population. We aimed to assess the incidence, timing and risk factors of pregnancy loss in cases with major fetal CHD, overall and according to cardiac diagnosis. This was a retrospective, population-level cohort study of fetuses and infants diagnosed with major CHD between 1997 and 2018 identified by the Utah Birth Defect Network (UBDN), excluding cases with termination of pregnancy and minor cardiovascular diagnoses (e.g. isolated aortic/pulmonary pathology and isolated septal defects). The incidence and timing of pregnancy loss were recorded, overall and according to CHD diagnosis, with further stratification based on presence of isolated CHD vs additional fetal diagnosis (genetic diagnosis and/or extracardiac malformation). Adjusted risk of pregnancy loss was calculated and risk factors were assessed using multivariable models for the overall cohort and prenatal diagnosis subgroup. Of 9351 UBDN cases with a cardiovascular code, 3251 cases with major CHD were identified, resulting in a study cohort of 3120 following exclusion of cases with pregnancy termination (n = 131). There were 2956 (94.7%) live births and 164 (5.3%) cases of pregnancy loss, which occurred at a median gestational age of 27.3 weeks. Of study cases, 1848 (59.2%) had isolated CHD and 1272 (40.8%) had an additional fetal diagnosis, including 736 (57.9%) with a genetic diagnosis and 536 (42.1%) with an extracardiac malformation. The observed incidence of pregnancy loss was highest in the presence of mitral stenosis (< 13.5%), hypoplastic left heart syndrome (HLHS) (10.7%), double-outlet right ventricle with normally related great vessels or not otherwise specified (10.5%) and Ebstein's anomaly (9.9%). The adjusted risk of pregnancy loss was 5.3% (95% CI, 3.7-7.6%) in the overall CHD population and 1.4% (95% CI, 0.9-2.3%) in cases with isolated CHD (adjusted risk ratio, 9.0 (95% CI, 6.0-13.0) and 2.0 (95% CI, 1.0-6.0), respectively, based on the general population risk of 0.6%). On multivariable analysis, variables associated with pregnancy loss in the overall CHD population included female fetal sex (adjusted odds ratio (aOR), 1.6 (95% CI, 1.1-2.3)), Hispanic ethnicity (aOR, 1.6 (95% CI, 1.0-2.5)), hydrops (aOR, 6.7 (95% CI, 4.3-10.5)) and additional fetal diagnosis (aOR, 6.3 (95% CI, 4.1-10)). On multivariable analysis of the prenatal diagnosis subgroup, years of maternal education (aOR, 1.2 (95% CI, 1.0-1.4)), presence of an additional fetal diagnosis (aOR, 2.7 (95% CI, 1.4-5.6)), atrioventricular valve regurgitation ≥ moderate (aOR, 3.6 (95% CI, 1.3-8.8)) and ventricular dysfunction (aOR, 3.8 (95% CI, 1.2-11.1)) were associated with pregnancy loss. Diagnostic groups associated with pregnancy loss were HLHS and variants (aOR, 3.0 (95% CI, 1.7-5.3)), other single ventricles (aOR, 2.4 (95% CI, 1.1-4.9)) and other (aOR, 0.1 (95% CI, 0-0.97)). Time-to-pregnancy-loss analysis demonstrated a steeper survival curve for cases with an additional fetal diagnosis, indicating a higher rate of pregnancy loss compared to cases with isolated CHD (P < 0.0001). The risk of pregnancy loss is higher in cases with major fetal CHD compared with the general population and varies according to CHD type and presence of additional fetal diagnoses. Improved understanding of the incidence, risk factors and timing of pregnancy loss in CHD cases should inform patient counseling, antenatal surveillance and delivery planning. © 2023 International Society of Ultrasound in Obstetrics and Gynecology.