RESUMEN
Smoking has become more prevalent in Turkey than it has in those of western countries during the past decade. This study was conducted to make parameter estimations on gender related smoking habits with the minimum of variance. Of the ninety-two researches related to smoking habits conducted from 1981 to 2003 in Turkey, 60 were deemed appropriate for the application of Meta analysis and Meta regression analysis. The proportions of men and women smoking cigarettes were 0.51 and 0.35, respectively. The proportion of men smoking cigarette in 1996 and the years before it was 0.52, and for women as 0.35. However, the figures for the years following 1996 were 0.41 for men, and 0.32 for women. In the results of the Dersimonian and Laird random effect model, the Odds Ratio, which shows the tendency of men to smoke compared to women, was found 1.894 for the period of 1981-2003. A heterogeneous distribution between the researches was apparent [Q=1560.91, P<0.001] as well as for Tau-square test [x2=0.55, z=6.29, P<0.001]. We propose that effective precautions should be considered, especially with regard to the introduction of laws to minimize the smoking habit for both sexes, with particular attention to women