Plasma clearance of creatinine was evaluated for assessment of glomerular filtration rate (GFR) in dogs. In 6 healthy dogs (Experiment 1), we determined 24-hour urine clearance of endogenous creatinine, plasma, and urine clearances of exogenous creatinine administered at 40, 80, and 160 mg/kg in a crossover design (linearity study), plasma iothalamate clearance, and plasma and urine clearances of 14C-inulin. In Experiment 2, plasma creatinine and iothalamate clearances were compared, and a linearity study was performed as for Experiment 1 in 6 dogs with surgically induced renal impairment. Experiment 3 compared plasma creatinine clearance with plasma iothalamate clearance before and 3 weeks after induction of moderate renal impairment in 6 dogs. Plasma creatinine clearances were calculated by both noncompartmental and compartmental analyses. In Experiment 1, plasma inulin clearance was higher (P < .001) than other clearance values. Plasma creatinine clearances at the 3 dose rates did not differ from urine inulin clearance and each other. In Experiment 2, plasma creatinine clearances were about 14% lower than plasma iothalamate clearance (P < .05). In Experiment 3, decreases in GFR assessed by plasma clearances of iothalamate and creatinine were similar. Renal failure decreased the daily endogenous input rate of creatinine by 25%. Limiting sampling strategies for optimizing GFR calculation were proposed, allowing an error lower than 6.5% with 4 blood samples. These results suggest that determination of plasma creatinine clearance by a noncompartmental approach offers a reliable, inexpensive, rapid, and convenient means of estimating GFR in routine practice.