The aim of this study was to determine the effect of the route of tobacco smoke exposure on urinary cotinine levels of infants. A cross-sectional analysis was done on 254 six-month-old infants. The infants were grouped according to the route of tobacco smoke exposure. The urinary cotinine/creatinine ratios were determined. Forty-nine percent (124/254) of mothers were smokers. Urinary cotinine levels in infants of smoking mothers were statistically significantly higher than levels in infants of non-smoking mothers. The highest mean cotinine/creatinine level was found in the breast-milk-exposed group. Linear regression analysis showed that maternal smoking increased urinary cotinine level by 541 times and breastfeeding increased it by 171 times, whereas early start of formula feeding decreased it by 63 times. Tobacco exposure by breastfeeding may be more harmful than other routes of exposure. Mothers should be encouraged to stop smoking during the breastfeeding period even if they avoid exposing their infants to passive tobacco smoke.