Experience with more than 150 pregnancies of women with systemic lupus erythematosus demonstrates that: many conventional measures of lupus activity, including complement, platelet count and urinary protein, are invalid during pregnancy; pregnancy does not cause lupus exacerbation; anti-phospholipid antibody is common and is closely associated with fetal loss, but is not the sole determinant factor of fetal loss; specific characteristics of anti-phospholipid antibody do not identify which antibody-positive women will have poor fetal outcome; prednisone therapy does not improve fetal prognosis; and neonatal lupus, diagnosed by rash and thrombocytopenia, is common but congenital heart block is rare.