Chromosomal deletion appears to be the earliest as well as the most frequent somatic genetic alteration during carcinogenesis. It inactivates a tumor suppressor gene in three ways, that is, revealing a gene mutation through loss of heterozygosity as proposed in the two-hit theory, inducing haploinsufficiency through quantitative hemizygous deletion and associated loss of expression, and truncating a genome by homozygous deletion. Whereas the two-hit theory has guided the isolation of many tumor suppressor genes, the haploinsufficiency hypothesis seems to be also useful in identifying target genes of chromosomal deletions, especially for the deletions detected by comparative genomic hybridization (CGH). At present, a number of chromosomal regions have been identified for their frequent deletions in prostate cancer, including 2q13-q33, 5q14-q23, 6q16-q22, 7q22-q32, 8p21-p22, 9p21-p22, 10q23-q24, 12p12-13, 13q14-q21, 16q22-24, and 18q21-q24. Strong candidate genes have been identified for some of these regions, including NKX3.1 from 8p21, PTEN from 10q23, p27/Kip1 from 12p13, and KLF5 from 13q21. In addition to their location in a region with frequent deletion, there are functional and/or genetic evidence supporting the candidacy of these genes. Thus far PTEN is the most frequently mutated gene in prostate cancer, and KLF5 showed the most frequent hemizygous deletion and loss of expression. A tumor suppressor role has been demonstrated for NKX3.1, PTEN, and p27/Kip1 in knockout mice models. Such genes are important targets of investigation for the development of biomarkers and therapeutic regimens.