Objective: To study the factors associated with an adverse pregnancy outcome in women with systemic lupus erythematosus (SLE). Methods: SLE women from LUMINA of Hispanic, African American and Caucasian ethnicity were studied. Adverse pregnancy outcome was a miscarriage or abortion (<20 weeks), a stillbirth (≥ 20) and/or a moderate to severe preterm-baby (<34 weeks); good outcome was either a mild preterm-baby (≥ 34 weeks) or a full-term baby [C-section or vaginal delivery (38-42 weeks)]. Pregnancies occurring after SLE diagnosis (TD) were included; pregnancy outcome was the unit of analyses. The relationship between selected variables and pregnancy outcomes was examined by univariable and multivariable analyses. Results: Adverse outcomes occurred in 63.7% of 102 pregnancies. In the univariable analyses, Texan Hispanic and African American ethnicities, fewer years of education, higher number of ACR criteria, renal involvement, glucocorticoid exposure and the maximum dose of glucocorticoids used prior to the pregnancy outcome were associated with an adverse pregnancy outcome. Renal involvement was independently associated with an adverse pregnancy outcome [Odds ratio (OR)=5.219 (95% Confidence Interval (CI) 1.416-19.239, p=0.0131] as were the maximum dose of glucocorticoids used prior to the pregnancy outcome (OR=1.028; CI:1.002-1.054; p=0.0315) and fewer years of education (OR=1.204; CI:1.006-1.472; p=0.0437). Ethnicity was not retained in the multivariable model. Conclusions: Renal involvement, the maximum dose of glucocorticoids used prior to pregnancy and fewer years of education were associated with adverse pregnancy outcomes. These data have implications for the management of women with lupus planning to become pregnant. © Copyright Clinical and Experimental Rheumatology 2008.