It is generally accepted that living in families where there are smokers, children are stressed not only by the harmful physical exposure to second-hand and third-hand tobacco smoke, but also by the negative models of the adult relatives' behaviour, as relatives who smoke can inspire children to imitate this behaviour, influencing attitudes towards, and early experiments with smoking. In this paper, some of the most important results about influence of family smoking on the effects of the anti-smoking educational programme "Non-smoking Is Normal" are described. The school-based programme was created by medical and educational specialists and targets children at the first level of primary schools (aged from 6 to 11 years). The data about interesting outcomes of the programme (knowledge, attitudes, behaviour) were collected by anonymous questionnaire, administered twice in each school year: one month before the complex of 5 lectures (pre-tests) and 4-5 months after the last lecture (post-tests). The sample of participants (860-910) was divided into four groups, according to the intervention and family backrounds: (1) programme children from smoking families "P-S"; (2) control children from smoking families "C-S"; (3) programme children from non-smoking families "P-NS"; (4) control children from non-smoking families "C-NS". The differences in the frequency of children's answers were analysed using the tests in statistic Epi Info software, version 6.04a (chi-square, Mantel Haenszel, Yates, Fisher). In the programme group, the number of children with smoking relatives was significantly higher than in the control group (80.1% vs. 73.0%, p < 0.01) as well as of those who reported frequent/daily exposure to secondhand smoke at homes and/or in cars (49.5% vs. 40.0%, p < 0.01). Smoking families significantly influenced the children's seeking of smoking friends (40% vs. 17%, p < 0.01). The programme has significantly increased the amount of knowledge about health risks of smoking. Both in the programme and control groups of children from non-smoking families, the frequency of critics of adults smoking was significantly higher all the time of the study (p < 0.05 and 0.01 resp.); however, the programme influenced children's opinions about smoking (criticism) only partially. Children's actual intentions about their smoking in the future was fully influenced by their smoking household environment: the number of "future no-smokers" has decreased in time and was significantly less frequent among children from smoking families (p < 0.01). The frequency of those willing to smoke significantly increased within the period between 3rd and 5th grades, both in the programme and the control groups (p < 0.01). An almost linear increase of active experimentations with cigarette smoking in follow-up monitoring was seen, trends of smoking children were steeper in groups from smoking families. The number of experiments with smoking was significantly lower in programme children of non-smoking parents only at the end of the study (p < 0.05). Despite of the effort to initiate parental participation on the primary prevention of smoking, we have confirmed that smoking in families decreased the efficacy of anti-smoking intervention targeted on young children at school age.