Background. The contribution of genetic factors to obsessive-compulsive (OC) symptoms has not been examined using a large population-based sample of adults. Furthermore, the extent to which there are qualitative and quantitative differences in genetic architecture between men and women with OC symptoms has not been elucidated. Method. We obtained the Young Adult Self Report Obsessive-Compulsive Scale (YASR-OCS) from a group of 5893 monozygotic (MZ) and dizygotic (DZ) twins, and 1304 additional siblings from the population-based Netherlands Twin Register. Structural equation modelling was used to decompose the variation in OC behaviour into genetic and environmental components and analyse quantitative and qualitative sex differences. Results. Familial resemblance was the same for DZ twins and non-twin siblings, which means that there was no evidence for a special twin environment. The same genetic risk factors for OC behaviour were expressed in men and women. Depending on the choice of fit index, we found small (39% for men and 50% for women) or no sex differences (47% for both men and women) in heritability. The remaining variance in liability was due to individual-specific environment. Conclusions. OC behaviour showed a moderate heritability. At most, small quantitative sex differences were found in the genetic architecture of OC behaviour, and no qualitative sex differences. © 2007 Cambridge University Press.