To examine patterns of interaction between cigarette smoking and genetic factors in relation to airways obstruction, cross sectional data were analysed on 1787 white non-patient adult participants in a genetic-epidemiological study of airways obstruction (AO), defined as one-second forced expiratory volume FEV1 less than 68% of forced vital capacity FVC. Interaction was examined between smoking and each of four factors previously found to be related to AO: alpha-1 antitrypsin (PiZ allele), ABO blood groups (A antigen), ABH non-secretor status, and first degree relationship to a COPD or lung cancer patient. Multiple linear regression was used to test for interaction and adjust mean FEV1 (as a per cent of FVC) and prevalence of AO for age, sex, socioeconomic status, coffee and alcohol intake. Statistical interaction was observed between smoking (measured in pack-years) and two genetic factors (presence of blood A antigen and the family history). At higher pack-year levels, those individuals with the A antigen or the family history, but especially those with both factors had a much lower mean FEV1/FVC % and a much higher prevalence of AO than expected based on a simple additive model. On the other hand, there was no interaction between smoking and PiZ allele, or smoking and ABH secretor status. The findings suggest a possible biological interaction between cigarette smoke and the airways of individuals with blood group A antigen and familial lung disease. The findings also emphasize the role of genetic-environmental interactions in chronic diseases of multifactorial aetiology.