The comet assay (single-cell gel electrophoresis, SCG) is widely accepted as an in vitro and in vivo genotoxicity test. Because of its demonstrated ability to detect various kinds of DNA damage and its ease of application, the technique is being increasingly used in human biomonitoring. However, the assessment of small genotoxic effects as typically obtained in biomonitoring may be limited by the different sources of assay variability and the lack of an optimal protocol with high sensitivity. To better characterize the suitability of the comet assay for biomonitoring, we are performing a comprehensive investigation on blood samples from smokers and non-smokers. Because tobacco smoke is a well-documented source of a variety of potentially mutagenic and carcinogenic compounds, smokers should be a suitable study group with relevant mutagen exposure. Here, we report our results for the first sample of 20 healthy male smokers and 20 healthy male non-smokers. Baseline and benzo[a]pyrene diolepoxide (BPDE)-induced effects were analysed by two investigators using two image analysis systems. The study was repeated within 4 months. Furthermore, the influence of a repair inhibitor (aphidicolin, APC) on baseline and BPDE-induced DNA damage was comparatively analysed. In all experiments, a reference standard (untreated V79 cells) was included to correct for assay variability. None of these approaches revealed significant differences between smokers and non-smokers. Although more data is needed for a final conclusion, this study indicates some limitations of the comet assay with regard to the detection of DNA damage induced by environmental mutagens in peripheral blood cells.