Smoking has been associated with increased incidence of rheumatoid arthritis (RA), joint damage and positive rheumatoid factor (RF). Here we report an analysis of the association between smoking and IgM, IgG and IgA RF in a cohort of non-rheumatic individuals participating in a prospective longitudinal study of the incidence and significance of elevated RF. From the initial cohort of nearly 14,000 randomly selected individuals aged 52-80 years, 109 RF-positive and 187 RF-negative non-rheumatic participants were recruited. All participants were tested for RF at least twice at an interval ranging from 4 to 13 years. Of the RF-negative participants 21.9% were active smokers compared to 34.1% of IgM RF-positive (p=0.035), 20.8% of IgG RF-positive (N.S.) and 34.4% of IgA RF-positive participants (p=0.047). Smoking was most prevalent (44.8%) amongst participants with elevation of both IgM and IgA RF (p=0.008), and smokers were also significantly more likely to have a persistent elevation of RF than non-smokers (p=0.024). These findings indicate that smoking may influence the immune system, leading to increased production of IgM and IgA RF.