Fifty-one infants and small children (14.7 +/- 7.2 mo) were studied to determine the MAC of halothane in O2 (n = 11) and in the presence of three different nitrous oxide (N2O) concentrations (25% [n = 13], 50% [n = 13], and 75% [n = 14]). In the three N2O groups, after randomly assigning patients to an N2O group, anesthesia was induced with halothane and N2O using a pediatric circle system. After endotracheal intubation, halothane and N2O end-expired concentrations were adjusted to predetermined concentrations. The initial halothane concentrations in each group were based on the assumption that each percent N2O reduced halothane concentrations by 0.01 vol % (assumed halothane MAC = 1.0 vol %). Based on the response of the preceding subject in each group, halothane concentrations were increased or decreased depending on whether the response was to move or not to move, respectively, in response to the surgical incision. The mean duration of constant end-tidal concentrations before skin incision was 10 min. End-tidal gases were sampled and measured from a separate distal sampling port of an endotracheal tube during controlled ventilation (Perkin-Elmer Mass Spectrometer). The MAC value for halothane in O2 was 0.94 +/- 0.08 vol % (mean +/- SD). The MAC values of halothane in the presence of 25%, 50%, and 75% N2O were 0.78 +/- 0.12 vol %, 0.44 +/- 0.10 vol %, and 0.29 +/- 0.06 vol %, respectively. All concentrations of N2O significantly reduced the MAC of halothane. A regression analysis through all four data points yielded a linear relationship (r2 = 0.87) with a predicted MAC for N2O of 105 vol %. Unlike halothane and isoflurane, the predicted MAC of N2O in infants and children is similar to that reported by others in adults. Similar to the results of clinical studies in adults, the contribution of N2O to halothane MAC in children is additive.