Interpersonal linkages were studied in 178 Hodgkin's disease patients, aged 60 years or younger, who lived in Western Australia between 1964 and 1975, and in their matched controls. Eighty-nine living subjects were interviewed about places and periods of residence, school attendance and employment, and possible linkages were computed based on concurrence of these events. Subjects were also shown the names of all patients and controls and asked to mark the names they recognized, giving details of acquaintanceships. The acquaintanceship method yielded more and the concurrence method fewer case-case links than expected. Little overlap occurred in linkages identified by the two methods. The acquaintanceship method is thought to be the more reliable. Risk factors suggested in the literature were also investigated. Increased risk of Hodgkin's disease in living patients was associated with being unmarried, being born outside Western Australia, smoking cigarettes, and having lived and worked on a farm and worked with animals. These effects did not explain the excess of case-case linkages found by the acquaintanceship method.