Following the ingenious prediction of Alfred Knudson in 1971, the first tumor suppressor gene, RB1, has been isolated. Its product, the RB1 protein, was found to play a major role in the control of the cell cycle. The loss of heterozygosity (LOH) technique, introduced by Cavenee and colleagues, was an important milestone toward the confirmation of Knudson's hypothesis and the identification of the gene. Subsequently, the LOH technique has provided important clues that have led to the discovery of other tumor suppressor genes. Most of them play important roles in the regulation of the cell cycle and/or of apoptosis. Circumstantial evidence suggests that still other and perhaps many unknown genes may participate in the protection of the organism against malignant growth. The numerous genome losses in tumors, detected by LOH, comparative genomic hybridization, and by cytogenetic techniques, support this possibility. The early work of one of us (G.K.), together with Henry Harris and Francis Wiener, had shown that the malignant phenotype can be suppressed by hybridizing malignant with low- or non-tumorigenic cells. However, analysis of this phenomenon failed to assign the inhibition of tumorigenicity to any particular gene. We have pursued the search for new tumor-antagonizing genes with two unconventional approaches, focusing on human chromosomal subband 3p21.3, a region frequently targeted by cytogenetically detectable deletions. We have detected four clusters of candidate tumor suppressor genes at 3p21.3 by a combination of deletion mapping and the "elimination test." These findings raise the question whether the number and variety of genes that may contribute to the defense against uncontrolled proliferation may have been underestimated.