BackgroundOver the last decade notable progress has been made in developed countries on monitoring smoking although experimenting with cigarettes and smoking in young people remains a serious public health problem. This paper reports a cross-sectional study at the beginning of the 3-year follow-up community study TA_BES. The aim was to study the prevalence of smoking in addition to determining predictive factors for when smoking commences in a representative population of 12-year-old first year compulsory secondary education students.MethodsTwenty-nine secondary schools (N = 29) from an area of Catalonia participated in the study. In these schools 2245 students answered a questionnaire to study the attitudes, behaviors, and tobacco consumption in the subject's surrounding circle and family in relation to smoking; carbon monoxide measurements were taken by means of co-oximetry on 2 different occasions. A smoker was defined as a student who had smoked daily or occasionally in the last 30 days. For non-smokers the criteria of not considering was set up for those who answered that in the future they would not be smokers and considering those who answered that they did not rule out becoming a smoker in the future.ResultsAmong the total 2245 students included in the analysis 157(7%) were classified as smokers. Among non-smokers we differentiated between those not considering smoking 1757 (78.3%) and those considering smoking 288 (12.8%).Age is among the factors related to commencing smoking. The risk of becoming a smoker increases 2.27 times/year. The influence of the group of friends with a very high risk for boys OR 149.5 and lower, albeit high, in girls OR 38.1. Tobacco consumption of parents produces different effects in young people. A smoking father does not produce alterations in the smoking behavior of young people. However having a smoking mother or former smoking is a risk factor for boys and a protective factor for girls.We detected a gradual risk of becoming a smoker by means of the co-oximetry test. A boy/girl with a test between 6 p.p.m and 10 p.p.m increased the probability of smoking by 2.29 and co-oximetry values > 10 p.p.m multiplied the risk 4 times over.ConclusionsResults indicate that the age of commencing smoking is maintained in spite of prevalence having decreased in the last few years. The risk factors identified should be used to involve families and the educational community by offering them tobacco weaning programmes.