The concentrations of lidocaine and mepivacaine were measured in stored cord blood samples from 200 deliveries. Newborn and obstetrical charts were reviewed for 100 deliveries for study of the association between cord blood concentrations, drug and route of recorded administration, Apgar scores at birth, and symptoms of possible toxicity. Detectable concentrations were commonly found (117/200), but toxic (3 micrograms/ml) levels were uncommon (4/200). However, recovery studies indicate these levels underestimate maximal exposure because both lidocaine and mepivacaine levels were shown to decrease in refrigerated stored blood. Although toxicity was seldom suspected, diagnostic accuracy was poor. The diagnosis was missed in the two patients with very high levels (5.0 and 9.0 micrograms/ml) of mepivacaine and levels measured in two suspected cases of toxicity were both low (0.4 microgram/ml). Associations between Apgar scores, cord levels, and route of administration (especially for lidocaine) were examined for 65 full-term uncomplicated deliveries. However, evaluation of these associations is problematic because dosages given, time or route of administration, and/or Apgar scores were often either not given or inconsistent with the clinical history. Detectable and potentially clinically significant local anesthetic drug concentrations were found in cord blood after all methods of administration, including local infiltration. These levels appear to underestimate the level of exposure because of instability of these anesthetics in stored cord blood samples. Local anesthetic toxicity appears to be difficult to detect clinically and may require cord blood level monitoring to detect.