We document the accuracy, efficacy, and safety of ultrasound in estimating reproductive characteristics of gray wolves (Canis lupus) in central Alaska. We examined 68 adult female wolves with ultrasound during late March and early April to diagnose pregnancy and litter size. Seventy-two percent were pregnant. We compared ultrasound diagnoses with postmortem embryo or placental scar counts in 14 females that died within 10 months of being examined by ultrasound; all ultrasound and postmortem examinations agreed in the diagnoses of pregnancy. Among 12 pregnant females, 6 agreed exactly in fetal count, 11 were within 1 fetus, and all were within 2 fetuses. In the postmortem sample we detected a decline in average placental scar color density between mid-September and mid-February. Radiocollared females were monitored from the air to estimate denning rates. Distance from the den declined as parturition approached, but few females localized near dens before parturition. Among 46 pregnant females diagnosed by ultrasound, 80.4% entered and remained at dens, 15.2% failed to enter dens, and 4.4% denned but abandoned the den within 1 week. None of the females diagnosed as nonpregnant entered dens. We present models of fetal growth from ultrasound measurements of embryonic vesicle diameters (EVD) or crown–rump length (CRL) of in utero fetuses. CRL was a better predictor of gestational age (r² = 0.92) than was EVD (r² = 0.79). We found no evidence that capture of females during the 2nd trimester of pregnancy affected denning or productivity.