The pharmacokinetics of intravenous theophylline were prospectively studied in 179 premature babies. Interrelated variables were analyzed for their influence on theophylline serum clearance. Gestational age, gender, duration of treatment, body weight, and Apgar scores were not found to correlate significantly with theophylline clearance. Weak but statistically significant correlations were found between serum clearance and postnatal (p less than 0.005) and postconceptional age (p less than 0.01). No significant difference in mean serum clearance (Cls) values was found between small-for-gestational-age (SGA) patients (Cls = 17.9 +/- 5.3 ml/kg/h) and appropriate-for-gestational-age (AGA) patients (Cls = 18.8 +/- 5.8 ml/kg/h). Conversely, asphyxiated patients had significantly lower mean clearance values than nonasphyxiated patients (16.4 +/- 5.3 ml/kg/h vs. 20.2 +/- 5.4 ml/kg/h, respectively, p less than 0.001). Volume of distribution for theophylline (n = 147) was 0.77 +/- 0.17 L/kg; there was no significant difference in distribution volumes between asphyxiated and nonasphyxiated patients or between SGA and AGA patients. Step-wise multiple regression analysis revealed postnatal age as the most important determinant of theophylline clearance among the variables analyzed (p less than 0.01). Postconceptional age had a statistically significant association with theophylline clearance in the entire group (n = 179, p less than 0.05). Duration of treatment had a small and statistically borderline effect (p less than 0.10) on theophylline clearance among nonasphyxiated infants when age factors were considered. Analysis of covariance confirmed the statistical effects of both postnatal age and asphyxia on theophylline serum clearance.