The objective of this study was to ascertain the complications and the efficacy of low-dose aspirin (LDA) and prednisone therapy in women with pregnancy loss and "lupus anticoagulants" (LAC). During the period 1985 to 1993, 255 patients with two or more pregnancy losses (RPL) were tested for LAC with an activated partial thromboplastin time (aPTT) and a tissue thromboplastin inhibition index (TTI, normal value < 1.3). The diagnosis of LAC was established if two TTI values were > or = 1.3 or if a prolonged aPTT was measured in the patient's plasma that did not correct to normal by 1:1 mixing with normal plasma. We excluded patients with RPL who had only anticardiolipin antibodies. We treated 28 pregnancies in 21 women with LDA/prednisone for RPL associated with LAC. Therapy with LDA/prednisone was initiated as soon as a viable pregnancy was diagnosed. Therapy was continued until delivery in all but one case. Prednisone dose was minimized by measuring TTI and aPTT every 2 weeks and adjusting the dosage to maintain a TTI < or = 1.2 and to correct the aPTT to less than 36 seconds. Among the 28 pregnancies there were four (14%) first-trimester spontaneous abortions and four (14%) second-trimester fetal deaths. Of 20 surviving neonates (72%), seven were delivered after 37 weeks and 13 before 37 weeks (mean 35.9 +/- 2.3 weeks, range 31.5 to 40.4 weeks). Pre-term premature rupture of membranes occurred in three pregnancies, hypertensive disorders in six, and four small-for-gestational-age neonates were delivered (two stillborn). Mean birth weight of 20 surviving neonates was 2736 +/- 763 gm (range 900 to 3920 gm). Mean daily prednisone dose in 20 live births was 24.1 +/- 8.5 (SD) mg (range 11.3 to 49.3 mg/day) with mean duration of LDA/prednisone therapy of 185 +/- 40 days (range 97 to 223 days). Maximum prednisone dose was 60 mg/day (mean 36.8 +/- 12.7 mg/day). Only one serious maternal complication of LDA/prednisone therapy was observed. One neonate had talipes equinovarus that resolved without surgical therapy. LDA/prednisone therapy seemed effective and reasonably well tolerated in this population. These findings should be confirmed in a prospective, controlled investigation if such a trial can be organized and performed.