Arrhythmogenic effects of aminophylline (theophylline ethylenediamine) during halothane anesthesia have been reported but have not been related to serum theophylline levels. This study was designed to determine the arrhythmogenicity of therapeutic and toxic serum theophylline levels during halothane anesthesia. The study consisted of three parts. In part 1 (induction) six dogs were anesthetized for 15 minutes with 1% halothane in air and then given intravenous aminophylline, 50 mg/kg. In part 2 (maintenance) eight dogs were anesthetized for 2 hours with 1% halothane and then given intravenous aminophylline, 10 mg/kg. In part 3, after four additional hours of steady-state 1% halothane anesthesia, additional intravenous aminophylline, 25 mg/kg, was given to these eight animals. Three of six dogs in part 1 had arrhythmias following aminophylline, with serum theophylline levels ranging from 48 to 66 mg/L. No dog in part 2 had arthythmias following the 10 mg/kg dose of aminophylline, with serum theophylline levels of 14 to 23 mg/L. Six of eight dogs in part 3 had arrhythmias shortly after aminophylline, 25 mg/kg, with serum theophylline levels of 36 to 72 mg/L. Aminophylline administration after prolonger 1% halothane anesthesia appears free from arrhythmogenic effects if serum theophylline levels remain near the therapeutic range (10 to 20 mg/L). Aminophylline administration resulting in high serum theophylline levels (above 36 mg/L) causes ventricular arrhythmias when aminophylline is given during induction or maintenance of 1% halothane anesthesia. Arrhythmias usually (89%) begin with 5 minutes of aminophylline administration, and these arrhythmias always resolve spontaneously within 2 minutes of onset.