In this study, we assessed the effects of exposure to 4 percent and 6 percent carboxyhemoglobin on ventricular arrhythmias in 41 subjects (nonsmokers) with documented coronary artery disease. We used a randomized, double-blind, crossover design. On day 1, a training session with no exposure, the baseline carboxyhemoglobin level was measured, and a supine bicycle exercise test was done. On days 2 through 4, subjects were exposed to room air, 100 parts per million (ppm)2 carbon monoxide (target, 4 percent blood carboxyhemoglobin), or 200 ppm carbon monoxide (target, 6 percent blood carboxyhemoglobin), and they then did a supine bicycle exercise test. Radionuclide ventriculography was performed at rest and during exercise. Ambulatory electrocardiogram recordings were made during the four consecutive days to determine the frequency of premature ventricular contractions at various intervals. The frequency of single premature ventricular contractions per hour during exercise was significantly greater on the 6 percent carboxyhemoglobin day than on the room air day (167.72 +/- 37.99 for 6 percent carboxyhemoglobin compared with 127.32 +/- 28.22 for room air, p = 0.03). The frequency of multiple premature ventricular contractions per hour was also significantly greater during exercise on the 6 percent carboxyhemoglobin day compared with the room air day (9.59 +/- 3.70 for the 6 percent carboxyhemoglobin day compared with 3.18 +/- 1.67 for the room air day, p = 0.02). Patients who developed increased arrhythmias during exercise on the 6 percent carboxyhemoglobin day were significantly older than those who had no increased arrhythmia, and, in addition, exercised longer and had a higher peak workload during exercise. No effect of carbon monoxide exposure was seen on the 4 percent carboxyhemoglobin day.