BACKGROUND Cancer in the sinonasal tract is rare, but persons who have been occupationally exposed to wood dust have a substantially increased risk. It has been estimated that approximately 3.6 million workers are exposed to inhalable wood dust in EU. In previous small studies of this cancer, ras mutations were suggested to be related to wood dust exposure, but these studies were too limited to detect statistically significant associations. METHODS We examined 174 cases of sinonasal cancer diagnosed in Denmark in the period from 1991 to 2001. To ensure uniformity, all histological diagnoses were carefully reviewed pathologically before inclusion. Paraffin embedded tumour samples from 58 adenocarcinomas, 109 squamous cell carcinomas and 7 other carcinomas were analysed for K-ras codon 12, 13 and 61 point mutations by restriction fragment length polymorphisms and direct sequencing. Information on occupational exposure to wood dust and to potential confounders was obtained from telephone interviews and from registry data. RESULTS Among the patients in this study, exposure to wood dust was associated with a 21-fold increased risk of having an adenocarcinoma than a squamous cell carcinoma compared to unexposed [OR = 21.0, CI = 8.0-55.0]. K-ras was mutated in 13% of the adenocarcinomas (seven patients) and in 1% of squamous cell carcinomas (one patient). Of these eight mutations, five mutations were located in the codon 12. The exact sequence change of remaining three could not be identified unambiguously. Among the five identified mutations, the G-->A transition was the most common, and it was present in tumour tissue from two wood dust exposed adenocarcinoma patients and one patient with unknown exposure. Previously published studies of sinonasal cancer also identify the GGT --> GAT transition as the most common and often related to wood dust exposure. CONCLUSIONS Patients exposed to wood dust seemed more likely to develop adenocarcinoma compared to squamous cell carcinomas. K-ras mutations were detected in 13% of adenocarcinomas. In this study and previously published studies of sinonasal cancer the found K-ras mutations, were almost exclusively G --> A transitions. In conclusion, our study, based on a large representative collection of human SNC tumours, indicates that K-ras mutations are relatively infrequent, and most commonly occur in adenocarcinomas. Wood dust exposure alone was not found to be explanatory for the G-->A mutations, but combination of exposure to tobacco, wood dust, and possibly other occupational agents may be a more likely explanation. Overall, the study suggests a limited role for K-ras mutations in development of sinonasal cancer.