Placental insufficiency is regarded as the primary factor contributing to late-term abortion and perinatal death of foals. Often when problems associated with late-term pregnancy in the horse are manifest the condition is well-advanced and therapeutic intervention may not be effective in rescuing the pregnancy. If a compromised pregnancy due to placental insufficiency could be identified early, the pregnancy might be sustained through medical intervention. Because the placenta is the sole source of circulating relaxin in the mare, we hypothesized that systemic relaxin may serve as a biomarker of placental function and fetal well-being and a predictor of pregnancy outcome at delivery. To test this hypothesis we monitored plasma relaxin in mares (light breeds) with normal and problematic pregnancies from clinical cases presented to the veterinary hospital and in pregnant mares experimentally inoculated with Streptococcus equi zooepidemicus to induce uterine infection. Upon establishment of placentitis, mares were assigned to different therapeutic strategies and responsiveness was monitored. Blood was collected during the third trimester of pregnancy, and relaxin content was determined using a homologous equine relaxin radioimmunoassay. The results reported here show a positive relationship between low circulating relaxin and poor pregnancy outcome in mares with compromised placental function. While relaxin may have value as a diagnostic assay for identifying mares with high-risk pregnancies associated with placental dysfunction, the variable results obtained from mares undergoing drug treatment for experimentally induced placentitis make it difficult to determine the reliability of relaxin for evaluating therapeutic efficacy.