The division of obsessive-compulsive symptoms (OCS) into specific factors is now widely accepted. However, the utility of these categories for genetic studies remains unclear, as studies examining their heritability have been inconsistent. Less attention has been paid to the possibility that clinically significant obsessionality is primarily determined by a "core" group of OCS that crosses the boundaries between symptom subgroups. The aim of this study is to determine whether such a core group exists, and to compare its heritability to that of the more traditionally derived symptom factors. We examined the properties and heritability of obsessive-compulsive symptoms in college students, medical students, and obsessive-compulsive disorder (OCD) families using the Leyton Obsessional Inventory. In each of the three samples, we identified a core group of symptoms that comprised a single unique construct and accounted for over 90% of the variation of the four more traditional symptom factors. This core construct was highly correlated with OCD in our families and had a heritability estimate of 0.19 when OCD was not included as a covariate and 0.49 when OCD was included as a covariate. In contrast, the four symptom factors were not heritable. There appears to be an underlying unidimensional component to obsessionality, both in non-clinical and clinical samples. This component, which is heritable, accounts for the majority of the variation of the more traditionally derived symptom factors in our sample, and is composed of OCS that are not specific to any of the symptom subgroups.