Six hundred one consecutive total joint arthroplasty patients were divided into three groups using three different catheterization protocols. Each group was evaluated for preoperative and postoperative urinary tract infections, number of catheterizations, and number of subsequent urinary tract manipulations. Group 1 (165 patients) was treated with "as needed" intermittent catheterization. One hundred six of these patients (64%) required a minimum of one catheterization with a total of 265 (mean 2.4 per patient) catheterizations. One post-operative urinary tract infection was recorded. Group 2 (295 patients) was treated with in-and-out catheterization on one occasion followed by anchoring of a closed drainage system if needed. One hundred eighty-one patients (61%) required catheterization, 124 (69%) of these patients required anchoring of a closed system. Two (0.69%) of these patients developed postoperative urinary tract infections. Group 3 (140 patients) were treated with intraoperative sterile anchoring of a closed drainage system which was maintained for 48 hours or less. Ten patients required in-and-out catheterizations (maximum one per patient) after system discontinuance. No urinary tract infections developed. Using Fisher's Exact test, no statistical difference in infection rates was found between the three groups. The group 3 protocol was felt to offer several patient conveniences.