The purpose of this study was to investigate whether smokers outside buildings with work-place smoking bans smoke "harder" than those smoking in social settings. An unobtrusive random observational study of smokers followed by structured interview was used, with 143 smokers taking smoking breaks outside their office buildings and 113 smokers in social settings. The main outcome measurements were number of puffs per cigarette and cigarette smoking duration. The mean number of puffs per cigarette for the office building group was 18.7% greater than that for the social settings group (10.7 +/- 3.2 vs. 8.7 +/- 2.7, t = 5.58, df = 253, p < 0.001); 74.8% of smokers outside offices took more than the mean number of puffs for the group compared to 42.5% of smokers in social settings (chi 2 df 1 = 26.31, p < 0.0001). Mean cigarette smoking duration was 30.4% shorter for the work-place group than the social settings group (3.9 +/- 1.2 minutes vs. 5.6 +/- 2.6 minutes). Of smokers outside offices, 55.2% had a cigarette smoking duration between 3 and 4.59 minutes, while 53.1% of smokers in social settings took > or = 5 minutes to smoke the observed cigarette (chi 2 df 2 = 31.55, p < 0.0001). Smokers who scored at the 75th percentile on the Fagerstrom Tolerance Scale took a mean 9.5 +/- 2.6 puffs per cigarette compared to 9.3 +/- 2.7 puffs by those who scored in the 25th percentile on the scale (t = 0.34, df = 145, p = 0.73). Regardless of degree of nicotine dependency, smokers leaving work-stations to smoke outside buildings smoked their cigarettes nearly 19% "harder" than cigarettes smoked in social settings. The individual and public health benefits of reduced smoking frequency engendered by work-place smoking bans may be lessened by policies which allow smokers to take smoking breaks.