To define the nature and magnitude of knee morbidity and to identify causative factors, three groups of active and retired workers were interviewed and medically examined. Participants in the survey were 112 carpet layers and floor layers (CF), who kneeled and used a knee kicker in their occupation; 42 tilesetters, terrazzolayers, mosaiclayers, and stonelayers (TTMS), who kneeled but did not use a knee kicker; and 243 millwrights, bricklayers, and decorators (MBD), who seldom kneeled and did not use a knee kicker. The subjects answered two questionnaires: one on the lifetime prevalence of seven knee conditions, use of the knee kicker or power carpet stretcher, nonoccupational knee injury, and personal characteristics, and another on ten standardized knee symptoms. In addition, physical and x-ray examinations were conducted on a subset of 108 of these workers to validate the questionnaire responses. Unconditional logistic regression was used to identify those personal and occupational characteristics that best predicted any of the seven knee conditions. The CF were found to report substantially more knee morbidity, particularly bursitis, needle aspiration of knee fluid, skin infection of the knee, and miscellaneous knee conditions, than either the general United States white male population or a blue collar working population of comparable age, sex, and race. The TTMS, however, were found to report a prevalence of arthritis above that of the MBD and comparable to that of the CF. The authors recommend that further research be carried out to characterize the nature and extent of knee disease in workers who incur chronic knee trauma and to develop an effective substitute for the knee kicker for stretching carpet.