The feasibility of renal arterial infusion of nonbiodegradable microspheres as a model of chronic renal disease in dogs was evaluated. Resin-coated, styrene-divinyl benzene copolymer microspheres were infused into the kidneys of healthy adult Beagles by direct injections of both renal arteries in a single surgical procedure. Injections of 25-micrometer diameter microspheres had minimal effect on either the clinical status or serum values of the dogs. Histologic examination revealed the majority of the microspheres lodged within the capillary beds of the glomeruli, and little change to the kidneys. However, injections of 50-micrometer diameter microspheres caused significant increases in serum concentrations of urea nitrogen and creatinine. Histologically, the larger microspheres obstructed afferent arterioles and small arteries, which caused diffuse glomerular necrosis and nephron damage. With doses ranging from 1 to 3 million microspheres/dog, a correlation between the quantity of microspheres injected and severity of renal damage was observed. The optimal dose for producing a model of moderate renal disease was determined to be 1.8 million microspheres/dog (0.9 million microspheres/kidney). During long-term studies, microsphere-injected dogs fed a moderately restricted protein ration remained relatively azotemic, compared with control dogs on the identical ration. During the 5-month postsurgical period, the serum urea nitrogen concentration averaged 18.41 +/- 1.59 mg/dl (mean +/- SE) for the microsphere-injected dogs vs 9.31 +/-0.38 for the control dogs (P < 0.001). Similarly, the mean serum creatinine value was significantly higher (P = 0.020) for the microsphere-injected dogs, compared with the controls (1.23 +/- 0.12 mg/dl vs 0.94 +/- 0.03). In addition, the difference in mean endogenous creatinine clearance rates was statistically significant (microsphere-injected 1.02 0.05 ml/min/kg, vs control 1.53 +/- 0.06, P < 0.001).