Arteriovenous shunting was quantitated with 20 plus or minus 10 mu 99mtechnetium human albumin microspheres in 12 patients with angiographically hypervascular renal cell carcinomas. Arteriovenous shunting was diagnosed angiographically when the renal vein was opacified during the arterial phase. With use of an angiographic catheter 2 to 3 mCi. 99mtechnetium human albumin microspheres were first injected into the renal artery directly supplying the tumor. Thereafter, 0.5 to 1 mCi. microspheres were injected into the inferior vena cava. Scintillation counts over the right lung were obtained immediately after each injection. The ratio between the 2 counts established the rate of tumor shunting after correction for dose difference and decay. Nine patients without angiographic evidence of arteriovenous shunting had mean shunt rates of 15 per cent, whereas 3 patients with angiographically visible shunting had significantly higher shunt rates of 57 per cent (p less than 0.001). It is concluded that angiography discovers only massive arteriovenous shunting in patients with renal cell carcinoma, whereas a fair degree of arteriovenous shunting may go undetected. The significance of minimal or moderate arteriovenous shunting in patients with renal cell carcinoma needs further investigation. The microsphere quantitation method can be applied routinely to angiography.