Introduction: The efficacy of tight glycemic control for the prevention of death and renal failure in the general diabetic population is well established. However, in diabetic renal-allograft recipients, the effect of different treatment strategies on outcomes is undetermined., Methods: We conducted a cohort study of 798 diabetic, renal-allograft recipients transplanted at the Medical University of Vienna between 1990 and 2004. We studied the influence of glucose parameters and diabetes treatment on mortality and graft loss. Marginal-structural models and multivariable Cox regression analysis were used to control for confounding., Results: Maximal glucose levels but not HbA1c were independently associated with mortality. Being in the highest quartile of maximal glucose increased the adjusted risk of death by a factor of 2.2 (P value for trend 0.009). Furthermore, in patients receiving insulin, the risk of death was increased 1.7-fold (95% confidence interval 0.9-3.1) compared with diet and 2.0-fold (95% confidence interval 1.1-3.7) compared with oral medication. Maximal glucose, HbA1c, or diabetes treatment did not influence death-censored functional graft survival., Discussion: In conclusion, maximal glucose levels and insulin treatment were independently associated with higher rates of mortality in our cohort of diabetic, renal-allograft recipients. However, graft survival was unaffected.