The familial aggregation of rheumatoid arthritis was examined to determine factors modifying the risk of rheumatoid arthritis in first degree relatives of 165 cases ascertained from January 1, 1987, through March 31, 1987, using the Saint Margaret Memorial Hospital Rheumatoid Arthritis Registry, Pittsburgh, Pennsylvania, without regard to previous information concerning the occurrence of rheumatoid arthritis among their family members. The reported affection status of first degree relatives, verified through a structured clinical evaluation, revealed a false-positive reporting rate for family members of 61%. In contrast, there were no false-negative cases detected. There were no differences in average family size or total number of years at risk between 135 simplex and 30 multiplex families; however, aggregation analysis revealed that only 18 of 30 confirmed multiplex families had significant excess risk of rheumatoid arthritis. Significant differences were found when probands from multiplex families were compared with those from simplex families with regard to female to male ratio for probands (1:1 in multiplex families vs. 3:1 in simplex families) and average age of onset for probands (41 years in multiplex families vs. 48 years in simplex families). The familial risk for rheumatoid arthritis was similar in parents (4.2%) and siblings (4.6%) and lowest for children (0.7%) of probands. The authors assert that the affection status of first degree relatives of patients with rheumatoid arthritis is often falsely reported as positive. The familiality of rheumatoid arthritis may be more accurately related to the sex and age at onset of the affected family member.