Peripheral nerve blocks are becoming increasingly popular for perioperative use as anesthetics and analgesics in small animals. This prospective study was performed to investigate the duration of motor and sensory blockade following use of bupivacaine for ultrasound-guided femoral and sciatic nerve blocks in dogs and to measure the plasma concentrations of bupivacaine that result from these procedures. Six dogs were anesthetized twice using a randomized cross-over design. At the first anesthetic, dogs were assigned to receive either an ultrasound-guided femoral nerve block or sciatic nerve block with 0.15 mL kg-1 of bupivacaine 0.5%. Two months later, the other nerve block was performed during a second anesthetic. At 5, 10, 15, 20, 30 and 60 minutes after injection, arterial blood samples were collected for laboratory measurement of bupivacaine. After 60 minutes, dogs were recovered from anesthesia. Starting at two hours post-injection, video-recordings of the dogs were made every two hours for 24 hours. The videos were randomized and the degree of motor and sensory blockade was evaluated using a three-point scoring system (0 = no effect, 1 = mild effect, 2 = complete blockade) by two blinded assessors. The median (range) times to full recovery from motor blockade were 11 (6-14) hours (femoral) and 12 (4-18) hours (sciatic), and 15 (10-18) hours (femoral) and 10 (4-12) hours (sciatic) for sensory blockade. There were no differences in the median times to functional recovery for the two techniques. Plasma concentrations of bupivacaine were no different following the blocks and were less than 0.78 μg mL-1 at all times. These results suggest that these ultrasound-guided nerve blocks do not result in potentially toxic systemic levels of local anesthetic and that their duration of action is useful for providing anesthesia and analgesia for pelvic limb procedures.