This study extends the market mapping concept introduced by Counts et al. (Counts, M.E., Hsu, F.S., Tewes, F.J., 2006. Development of a commercial cigarette "market map" comparison methodology for evaluating new or non-conventional cigarettes. Regul. Toxicol. Pharmacol. 46, 225-242) to include both temporal cigarette and testing variation and also machine smoking with more intense puffing parameters, as defined by the Massachusetts Department of Public Health (MDPH). The study was conducted over a two year period and involved a total of 23 different commercial cigarette brands from the U.S. marketplace. Market mapping prediction intervals were developed for 40 mainstream cigarette smoke constituents and the potential utility of the market map as a comparison tool for new brands was demonstrated. The over-time character of the data allowed for the variance structure of the smoke constituents to be more completely characterized than is possible with one-time sample data. The variance was partitioned among brand-to-brand differences, temporal differences, and the remaining residual variation using a mixed random and fixed effects model. It was shown that a conventional weighted least squares model typically gave similar prediction intervals to those of the more complicated mixed model. For most constituents there was less difference in the prediction intervals calculated from over-time samples and those calculated from one-time samples than had been anticipated. One-time sample maps may be adequate for many purposes if the user is aware of their limitations. Cigarette tobacco fillers were analyzed for nitrate, nicotine, tobacco-specific nitrosamines, ammonia, chlorogenic acid, and reducing sugars. The filler information was used to improve predicting relationships for several of the smoke constituents, and it was concluded that the effects of filler chemistry on smoke chemistry were partial explanations of the observed brand-to-brand variation.