I t is not known whether mutations in the PKD1 gene cause autosomal dominant polycystic kidney disease (PKD) by an activating (gain-of-function) or an inactivating (loss-of-function) model. We analyzed DNA from cyst epithelial cells for loss of heterozygosity (LOH) in the PKD1 region of chromosome 16p13 using microsatellite markers. 29 cysts from four patients were studied. Five cysts from three patients had chromosome 16p13 LOH. Four of the cysts had loss of two chromosome 16p13 markers that flank the PKD1 gene. In two patients, microsatellite analysis of family members was consistent with loss of the wild-type copy of PKD1 in the cysts. In the third patient, 16p13 LOH was detected in three separate cysts, all of which showed loss of the same alleles. Chromosome 3p21 LOH was detected in one cyst. No LOH was detected in four other genomic regions. These results demonstrate that some renal cyst epithelial cells exhibit clonal chromosomal abnormalities with loss of the wild-type copy of PKD1. This supports a loss-of-function model for autosomal dominant PKD, with a germline mutation inactivating one copy of PKD1 and somatic mutation or deletion inactivating the remaining wild-type copy.