A preliminary study of maternal smokeless tobacco use, mostly oral applications of burnt tobacco or 'mishri', in pregnancy showed 65 of 178 singleton liveborns occurred to users and 113 to non-users in Bombay, India. Eighty-three newborns, 42 to maternal tobacco users and 41 to non-users were < 2.5 kg birth weight, i.e. low birth weight (LBW; odds ratio 3.2; confidence interval 1.5-6.9; P < 0.001). Stratifying by gender yielded odds ratios of 1.6 (P > 0.1, NS) for male and 6.96 (confidence interval 2.5-19.4, P < 0.0005), for female newborns compared to normal birthweight boys and girls, respectively. Male:female newborns were 80.6:100 in maternal tobacco users compared to 105.5:100 in non-users. Defining LBW as < 2.0 kg yielded an odds ratio of 5.4 (confidence interval 1.8-15.2, P < 0.005) in maternal tobacco users' offspring. For babies weighing 2-2.5 kg at birth it was 2.76 (confidence interval 1.4-5.5, P < 0.01). Maternal use of 'mishri' tobacco in pregnancy may be associated with (1) the offsprings' low birth weight, (2) low birth weights in girls more than in boys; (3) decreased male:female ratio in live newborns, and (4) low birth weight of < 2.0 kg more than of 2-2.5 kg. Studies are needed to substantiate these findings. Gender differences in outcome suggest the in utero effect of maternal smokeless tobacco use on male and female fetuses may differ.