Background It can be assumed that resettlers (ethnic German immigrants from the Former Soviet Union) show similar smoking patterns as persons in their countries of origin at the time of migration. We analysed how the smoking prevalence among resettlers differs from that among the general population of Germany and whether the prevalence differs between groups with increasing duration of stay. Methods To estimate the smoking prevalence we used the scientific-use-file (n = 477,239) of the German 2005 microcensus, an annual census representing 1% of all German households. Participation in the microcensus is obligatory (unit-nonresponse <7%). We stratified the prevalence of smoking among resettlers and the comparison group (population of Germany without resettlers) by age, sex, educational level and duration of stay. In total, 14,373 (3% of the total) persons were identified as resettlers. Results Female resettlers with short duration of stay had a significantly lower smoking prevalence than women in the comparison group. With increasing duration of stay their smoking prevalence appears to converge to that of the comparison group (e.g.: high educational level, age group 25-44 years: short duration of stay 15%, long duration of stay 24%, comparison group 28%). In contrast, the smoking prevalence among male resettlers with short duration of stay was significantly higher than that among men in the comparison group, but also with a trend towards converging (e.g.: high educational level, age group 25-44 years: short duration of stay 44%, long duration of stay 35%, comparison group 36%). Except for female resettlers with short duration of stay, the participants with low educational level had on average a higher smoking prevalence than those with a high educational level. Conclusions This is the first study estimating the smoking prevalence among resettlers by duration of stay. The results support the hypothesis that resettlers brought different smoking habits from their countries of origin shortly after migration. The observed convergence of the smoking habits with increasing duration of stay is in line with the hypothesis of migration as 'health transition'. However, due to the cross-sectional design of the study, further research is needed to confirm these findings.