A recent paper1 stated that maternal antibody to the N-methyl-D-aspartate receptor (NMDAR) causes death of female fetuses of BALB/c mice, leading to offspring male/female (m/f) ratios greater than the normal ratio of 1.0. Since about 40% of patients with systemic lupus erythematosus (SLE) have anti-NMDAR, if the mouse model is predictive, m/f ratios in human SLE pregnancies might exceed the United States normal ratio of 1.048. Our pregnancy studies offer an opportunity to test this prediction. PROMISSE (Predictors of Pregnancy Outcome: Biomarkers in Antiphospholipid Syndrome and Systemic Lupus Erythematosus) is a multicenter, prospective study of pregnancies of mothers with lupus;2 the Research Registry for Neonatal Lupus (RRNL) is a registry of families in which the mother has anti-SSA/Ro and/or SSB/La antibodies and at least one child with neonatal lupus. RRNL includes both affected and unaffected children.3 The table displays m/f ratios of liveborn children of RRNL and PROMISSE, stratified by maternal serology and maternal and fetal diagnosis, and compared to the PROMISSE control group of normal women and to the United States norm. (PROMISSE patients with antiphospholipid antibody were excluded). Ratios are similar if stillborn children are counted. The m/f ratios are 1.0 for children of mothers with negative tests for anti-NMDAR and <1.0 for those with positive tests (p = 0.36). Because anti-DNA antibodies often coexist with anti-NMDAR antibodies, we compared m/f ratios in anti-DNA positive and negative patients and found no difference (p = 0.30). M/f ratios of patients with and without pregnancy complications did not differ (p = 0.34). Thus our human data do not confirm the mouse experiments. Table Male/female ratios of liveborn children in RRNL and PROMISSE studies, stratified by maternal or fetal diagnosis, fetal sex, and maternal serology. Inclusion of stillborn children does not change the ratios. See text for abbreviations. Prior data on human SLE fetal sex ratios are weak: in a national birth registry containing 68,600 women, conclusions were based on the 42 women who had a diagnostic code identifying them has having lupus.4 A summary of published case series (diagnostic criteria unspecified) reported studies ranging in size between 14 and 249 children; the four largest series, constituting half the infants, had a combined m/f ratio of 1.07, 5 Our prospective studies, having rigorous definition of maternal diagnosis and fetal outcome, are more likely to be correct, and the mouse-based hypothesis of male predominance of liveborn children incorrect. Our data reemphasize that the mouse models do not invariably predict human outcomes.