The renal sinus is the fatty compartment located within the confines of the kidney not delineated from the renal cortex by a fibrous capsule. Because it contains numerous veins and lymphatics, invasion into this compartment may permit dissemination of a tumor otherwise regarded as renal-limited. Thirty-one consecutive renal carcinomas were studied: 22 clear cell renal cell carcinomas (3 multilocular cystic renal cell carcinomas), 4 chromophobe renal carcinomas, and 5 papillary renal carcinomas. The entire interface between the neoplasm and the sinus was embedded. Seventeen carcinomas did not invade the renal sinus and 16 were pT1 or pT2 tumors. Fourteen carcinomas, 13 clear cell renal cell carcinoma and one chromophobe renal carcinoma, invaded the renal sinus fat, and 9 of 14 invaded the lumen of renal sinus veins (all clear cell renal carcinomas). Although 14 of 22 clear cell renal carcinomas appeared to be renal limited pT1 and pT2 cancers, 6 of 14 carcinomas invaded sinus fat and 4 invaded into the lumen of renal sinus veins. Compared with the nine sinus-negative clear cell renal cell carcinomas, the 13 sinus-positive cancers were larger, exhibited more frequent renal capsule and renal vein involvement, and had higher nuclear grades. Renal sinus invasion was most common in clear cell renal cell carcinomas but was uncommon (one in 12) in 3 more indolent renal cell carcinomas: multilocular cystic renal cell carcinoma, chromophobe renal carcinoma, and papillary renal carcinoma. The follow-up period was short (1-17 months), but metastases developed in four of 31 cases. In three cases with metastases, carcinoma had involved the lumen of sinus veins but not the main renal vein, although two of three had also invaded through the renal capsule. This study shows that in carcinomas which appear to be renal limited (pT1/pT2), seven of 23 (30.4%) had invaded sinus fat and four of 23 (17.4%) had invaded sinus veins. We conclude that renal sinus invasion, especially sinus vein invasion, could identify a patient at risk for metastases even in a putative renal limited tumor, and suggest that all cases be examined for this feature. Renal sinus invasion merits further investigation to establish its prognostic importance and possible incorporation into future revisions of the TNM staging system for renal cell carcinomas.