T he fa mil ial na ture of ob ses sive-com pul sive dis or der (OCD) has been de scribed since the 1930s (1-14), but the sub ject re mains con tro ver sial. While some early fam ily studies re ported find ings that sup port the fa mil ial na ture of OCD (1-4,8), oth ers found no in crease in the rates of OCD among first-degree rela tives (5-7). Many ear lier stud ies are, however, dif fi cult to in ter pret be cause of dif fer ences in di ag nos tic cri te ria and meth ods of as sess ment. Most stud ies did not directly in ter view rela tives and lacked con trol groups.In some re cent stud ies, the meth odo logi cal short com ings of the ear lier stud ies have been ad dressed by di rectly in ter viewing rela tives and em ploy ing stan dard di ag nos tic cri te ria (9-16). Three of the stud ies (9-11) re ported a mor bid ity risk of 13% to 36% in fami lies hav ing chil dren with OCD. A ma jor meth odo logi cal weak ness of all 3 stud ies was the lack of a com pari son group. The study by Pauls and oth ers (14), which in cluded a com pari son group, re ported a higher risk for OCD (20%) among the rela tives of pro bands with early-onset OCD (less than age 18 years) than among those with lateonset OCD (11%), whilein rela tives of the psy chi at ri cally unaf fected com pari son group, the rate was 1.9%. In ter est ingly, 2 stud ies found evi dence for famili al ity only in rela tives of early-onset pro bands (12,16). All these stud ies strongly support the pos si ble fa mil ial com po nent for the ex pres sion of OCD, par ticu larly child hood-and adolescent-onset OCD.Ad di tional sup port for the role of fa mil ial fac tors in the expres sion of CD comes from the find ing of high rates of OCD among rela tives of pro bands with Tourette syn drome (TS) (15,(17)(18)(19). Simi larly, high rates of tic dis or ders and TS have been re ported among rela tives of OCD pro bands (11,14), sug gest ing that some forms of OCD may be etio logi cally (and per haps ge neti cally) re lated to TS.While many stud ies re port in creased rates of OCD in rela tives of OCD pro bands (9-12,14-16), find ings of some stud ies do not sup port the fa mil ial na ture of OCD (13,20). The study by Black and oth ers (13) found no evi dence that OCD was fa milial, and they con cluded that an anxi ety dia the sis rather than than OCD per se may be in her ited. A fam ily study from our cen tre (20), which ex am ined 135 adult first-degree rela tives