In a controlled, prospective study of 44 consecutive women with idiopathic habitual abortion, only 5% had symptoms of rheumatic disease. Patients did not differ from control subjects in the frequency of positive results on tests for antinuclear antibody or anti-doublestranded DNA. Levels of C3 and C4 were higher in the habitual aborters. No patients had anti-Ro. The antiphospholipid antibody results were analyzed using 2 methods: the frequency of antiphospholipid antibodies was 9% by lupus anticoagulant using the Russell viper venom time (95% confidehce interval 22-2.5). and 11% by anticardiolipin antibody assay (95% confidence interval 25-3.7), which was not significantly different from that in control subjects. However, the mean levels in the aborters (although within the normal range) were Submitted for publication August 18, 1986; accepted in revised form January 7, 1987. significantly higher than those in control subjects for anti-double-stranded DNA (P = 0.004), lupus anticoagulant (by Russell viper venom time; P = 0.09), and anticardiolipin antibody (P = 0.0007), when examined by multiple linear regression analysis corrected for age and concurrent pregnancy. Of the 3 patients with antiphospholipid antibodies and subsequent successful pregnancies, only 1 was treated with prednisone and aspirin. We conclude that, in the majority of women, subclinical lupus, anti-Ro, the lupus anticoagulant, and anticardiolipin antibodies are not associated with idiopathic habitual abortion.Women with systemic lupus erythematosus (SLE) have an increased frequency of spontaneous abortions (1-4), but most women who habitually abort do not meet criteria for the diagnosis of SLE (5,6). No previous study has examined the frequency of laboratory findings that are generally associated with lupus (antinuclear antibodies [ANA], anti-Ro, lupus anticoagulant [LAC], and anticardiolipin [ACL] antibody among women with idiopathic habitual abortion ( 3 or more consecutive spontaneous abortions) compared with the frequency among normal controls. In the absence of controlled trials, it has been proposed that ANA, anti-Ro, LAC, and ACL, antibodies play a causal role in idiopathic habitual abortion and that treatment of the problems indicated by these laboratory findings may be necessary (7-10).We now report the frequency of ANA, antidouble-stranded DNA (dsDNA), anti-Sm, anti-RNP, anticentromere antibody, anti-Ro, anti-La, cornple-