The aim of this study was to examine the association between uterine artery Doppler velocimetry discordance and perinatal outcome, specifically in pregnancies complicated by diabetes. We evaluated 265 women with singleton pregnancies complicated by diabetes who underwent Doppler ultrasonographic examinations of the right and left uterine arteries within 1 week before delivery. The absolute difference between the right and left uterine arteries was computed after measuring the uterine artery systolic-diastolic ratio. Adverse outcome was defined as still-birth, intrauterine growth restriction, delivery before 37 weeks' gestation, or cesarean delivery for fetal risk. The discordance between right and left uterine artery systolic-diastolic ratios ranged from 0 to 2.3, with a mean of 0.39 +/- 0.36 and a median of 0.30. The discordance was significantly larger in the 63 pregnancies with adverse outcome than in those with good outcome (0.48 versus 0.36, P = 0.018). Among the women with large uterine artery S/D ratio differences (> or = 0.60), a cesarean delivery for fetal risk was three times more likely (21.5% versus 7.5%, P = 0.002). In diabetic women with chronic hypertension (n = 36), the discordance was significantly larger than in the 201 normotensive women (0.54 versus 0.35, P = 0.001); yet for this subgroup uterine artery S/D ratio discordance was not predictive of adverse outcome. In conclusion, although considerable overlap in discordance exists between the good and adverse outcome groups, the uterine artery S/D ratio discordance added prognostic information on perinatal outcome for normotensive women with diabetes. The predictive value is independent of White's classification, third trimester glycemic control, sex of the infant, and umbilical artery Doppler waveform data.