In his introductory chapter of the Mutation Research special issue on 'Genetic Instability and Aging', the late Bernard Strehler provided some historical perspectives on the long-standing hypothesis that aging is primarily caused by changes in the genome of somatic cells (Strehler, 1995, Mutat. Res. 338 (1995) 3). Based on his own findings of a loss of ribosomal RNA gene copies in postmitotic tissues of dogs as well as humans during aging, his main conclusion was that deletional mutations are more likely than point mutations to be a main causal factor in aging. To directly assess the levels of different types of spontaneous mutations in organs and tissues during aging, we have used a mouse model harboring a chromosomally integrated cluster of lacZ-containing plasmids that can be recovered and analyzed in Escherichia coli. Our results indicate the accumulation of mutations in some but not all organs of the mouse with significant differences in mutational spectra. In addition to point mutations, genome rearrangements involving up to 66 Mb of genomic DNA appeared to be a major component of the mutational spectra. Physical characterization of the breakpoints of such rearrangements indicated their possible origin by erroneous, non-homologous DNA double-strand break repair. Based on their increased occurrence during aging in some tissues and their often very large size, we have designed a model for an aging tissue in terms of a cellular mosaic with a gradual increase in genome rearrangements that leads to functional senescence, neoplastic transformation or death of individual cells by disrupting nuclear architecture and patterns of gene regulation.