Reports of smoking cessation studies often claim that many relapsed subjects reduce their smoking. We investigated the smoking habits of relapsers 1 year after quitting in a smoking cessation trial using nicotine or placebo patches. All 289 participants in that study were summoned to a 1-year follow-up visit--148 (57%) of 259 relapsers attended, as did all 30 sustained abstainers. Fewer than 1% of the subjects had quit spontaneously after the primary relapse. Daily cigarette consumption, standard nicotine yield per cigarette, saliva cotinine concentration, expired carbon monoxide level and two nicotine dependency scales were assessed at entry and at the 1-year follow-up. In five of these six smoking-related characteristics, there was a small but significant mean reduction of 7%-27%. A significant weight gain of 0.5 +/- 2.9 kg (mean +/- SD) was recorded in the relapsers compared with 4.8 +/- 4.2 kg for abstainers (p less than 0.001). It is concluded that smoking habits in relapsers are relatively unchanged, and thus the most important outcome measure in smoking cessation trials is abstinent subjects.