Background. Pre-emptive kidney transplants have not been favored in some centers because of concern about possible increased noncompliance and allegedly inferior long-term results. We analyzed our experience with pre-emptive kidney transplants to determine whether such concerns are justified. Patients and Methods. Between January 1, 1984, and June 30, 1998, we performed 1849 adult primary kidney transplants: 385 pre-emptive (recipients not undergoing dialysis, ND) and 1464 non-pre-emptive (recipients undergoing dialysis, D). Results were subdivided by donor source: cadaver (CAD) and living donor (LD). ND recipients tended to be younger, but otherwise, the two groups were similar. Posttransplantation quality of life in recipients was evaluated using the nationally standardized Short Form Health Survey (SF-36). The posttransplantation employment status of the recipients was also evaluated. Results. The patient survival rate 5 years posttransplantation was significantly better for ND (vs. D) recipients for both CAD (92.6% vs. 76.6%, P=0.001) and LD (93.3% vs. 89.5%, P=0.02) transplants. The 5-year patient survival rate was significantly higher for ND recipients compared with recipients undergoing dialysis for < 1, 1-2, and > 2 years pretransplantation for both CAD (P=0.0005) and LD (P=0.0001) transplants. The graft survival rate 5 years posttransplantation was similar between ND and D recipients for CAD transplants, but significantly better for ND (vs. D) recipients of LD transplants (92.3% vs. 84.8%, P=0.006). For CAD transplants, the 5-year graft survival rate was not different when ND recipients were compared with recipients undergoing dialysis for < 1, 1-2, and > 2 years pretransplantation; for LD transplants it was significantly higher for ND recipients compared with recipients undergoing dialysis for < 1, 1-2, and > 2 years pretransplantation (P=0.04). The incidence of acute and chronic rejection was no different between ND and D recipients for either CAD or LD transplants, and it was also not affected by the pretransplantation time undergoing dialysis. Graft loss secondary to the recipient's discontinuation of immunosuppressive therapy (a crude estimate of compliance) was similar between ND and D recipients. Five years posttransplantation, the SF-36 scores regarding the recipient's quality of life and the employment status were similar for ND compared with D recipients, regardless of donor source. Conclusions. ND recipients do not seem to have higher rates of noncompliance than D recipients. Resuits for ND recipients seem to be superior than for D recipients, supporting the contention that renal failure patients should, if possible, undergo transplantation before dialysis.