1. Twelve dogs under light anaesthesia were infused with creatinine, lysozyme and other proteins, and the femoral arterial and urinary concentrations of the infused substances were measured. In some experiments renal venous blood was withdrawn through a catheter inserted by laparotomy. 2. The glomerular permeability to lysozyme was estimated to be 38% of the permeability to creatinine, but there was wide variation and the standard deviation was 11%. 3. The maximum tubular reabsorptive capacity for lysozyme was estimated to be 1·0 mg/100 ml glomerular filtrate, SD 0·44 mg/100 ml. 4. The renal threshold for lysozyme was about 1·0 mg/100 ml. 5. There was evidence supporting the hypothesis that reabsorbed lysozyme is catabolized and no evidence of reabsorption of intact lysozyme into the blood stream. 6. Infusion of myglobin, Bence Jones protein and ovalbumin had no effect on the excretion of endogenous or exogenous (infused) lysozyme. 7. Lysozyme infusions had no conclusive effect on the excretion of (infused) Bence Jones protein or ovalbumin. 8. Corrections, based on the recovery in vitro of lysozyme activity from dog serum and urine, do not significantly alter the results.