Increasing amount of genetic data on nicotine dependence (ND) is available in the literature, sometimes extremely large population size is reported but the study design is not always consequent. Phenotypic measures can vary from a simple 6-item self-rating scale to breath CO or serum cotinine level test but in genetic investigations this is not sophisticated; moreover the population stratification is also usually ignored. In contrast, possibly because of the strict traditions of pharmacological investigations, pharmacogenomic studies on smoking cessation therapy use more reliable phenotypic measures with high quality design consequently involving fewer participants. In spite of the heavy epidemiological data on smoking in Hungary, genetic background of heavy smoking is still not studied in this population. In this review we sum up the most important, replicated results but we also provide some critical remarks about the methodological shortcomings of these studies. Keeping in mind the value of large scale population ND association studies we would also like to emphasize that the clinical implementation of studies with larger samples but with weaker methodology and statistical analyses is limited. Similar to many other psychiatric disorders, ND is a multifactorial condition, therefore the measure of genetic effects requires a more complex study design.