Reduced body weight due to smoking may be an important factor inhibiting smoking cessation and promoting relapse after cessation in some smokers. It is popularly believed that smoking decreases body weight by suppressing appetite. However, cross-sectional studies show that, despite their lower body weights, smokers do not eat less than non-smokers or ex-smokers and, in fact, tend to eat slightly more. Similarly, laboratory studies show no acute effects of smoking or nicotine intake via other means on caloric intake in smokers, although intake of non-smokers may be reduced after nicotine. In contrast, longitudinal studies show that eating consistently increases in the first weeks after stopping smoking, but may recede to pre-cessation levels with longer-term abstinence, while resumption of smoking after cessation is accompanied by a reduction in eating. A similar pattern of results is seen for self-reported hunger and some, but not all, constituents of diet. Thus, there appear to be no acute or chronic effects of smoking on eating in smokers maintaining regular smoking, but changes in eating are observed concomitant with changes in smoking status (i.e. cessation or relapse). Although tolerance to anorectic effects of nicotine is one potential explanation, these findings may be more parsimoniously explained by viewing changes in eating due to smoking as secondary to an alteration in the set point around which body weight is regulated. According to this notion, cessation is accompanied by increased eating only until a new, higher body weight set point is reached, while relapse (and perhaps initiation of smoking) decreases eating only until a lower set point is reached. Implications of these findings and a set point explanation for them are discussed.