We evaluated the Jelliffe and the Cockcroft and Gault methods of estimating creatinine clearance in elderly nursing home patients (65 years) with chronic indwelling urethral catheters. Although these relationships have been prospectively validated in hospitalized and ambulatory populations previously, we found that they produced poor estimates of creatinine clearance in this patient group. For the Jelliffe method, 11 of 19 estimates were 20% or greater from the measured value; 10 of 11 poor estimates were high. The Cockcroft-Gault method had 10 of 19 estimates 20% or greater from the measured value with 8 of 10 of the poor estimates being high. An altered relationship between body weight, muscle mass, and daily creatinine production is the most likely explanation for the bias in these creatinine-clearance estimations. New estimates of the relationships between age, weight, serum creatinine, and creatinine clearance need to be developed for this population.