Tobacco smoking is associated with many health risks, all of which are decreased by smoking cessation. Tobacco companies are marketing novel products (e.g., Phillip Morris' Accord, R.J. Reynolds' Eclipse) intended to reduce these risks. Evaluation of these products is necessary to determine if they increase, decrease, or fail to change smoking's health risks. This study examines the acute effects of the Accord system on cigarette smokers; the study extends a previous preliminary evaluation by controlling for the system's eight-puff/cigarette limit, smokers' brand preference, and the act of smoking. Smokers of light or ultra-light cigarette brands (> or = 10 cigarettes/day, 16 men and 16 women) participated in this within-subjects protocol. Prior to each of the four, Latin-square ordered, approximately 2.5-h sessions, subjects abstained from smoking for at least 8 h. In each session, subjects smoked eight puffs, at 30-min intervals, from either their own brand of cigarettes, a marketed brand that was not their own, de-nicotinized tobacco cigarettes, or the novel smoking system. Subjective and physiological effects were assessed regularly. Results confirmed previous observations that the withdrawal suppression, CO intake, and tachycardia produced by the novel system are all less than that produced by normally marketed cigarette brands, even controlling for the system's eight-puff limitation and for brand preference. Incomplete withdrawal suppression may increase smoking frequency, thus potentially offsetting any decreased risks associated with Accord use. Laboratory studies of the acute effects of novel smoking systems are an important part of a comprehensive evaluation program.