Abstract Single-cell sequencing for analyzing DNA mutations across the genome in somatic tissues is critically important for studying development, cancer and aging. However, current procedures are prone to artifacts and to date a reliable protocol for single-cell somatic mutation analysis remains to be developed. We address the two largest sources of artifacts, i.e., DNA denaturation-related cytosine deamination and allelic bias-driven whole genome amplification errors. We first reconfigured multiple displacement amplification (MDA) into an efficient protocol for whole genome amplification of single cells without cytosine deamination artifacts, i.e., Single Cell MDA (SCMDA). We then developed a new single-cell SNV caller (SCcaller) that distinguishes real somatic mutations and amplification errors by utilizing a SNP-based localized estimate of allelic amplification bias. The procedure was validated by comparing SCMDA-amplified single cells with unamplified clones derived from single cells from the same population. Using this highly accurate single-cell whole-genome sequencing method we analyzed human B lymphocytes from donors varying in age from birth to over 100 year, studying both genome distribution and functional impact of base substitution mutations. Mutations per cell were found to increase with age from less than 500 in cord blood to over 3,000 in cells from individuals over 100. While overall mutations were randomly distributed across the genome with all chromosomes equally affected, 24 hotspot regions were identified, five of which were part of immunoglobulin variable regions subject to somatic hypermutation. Age-related mutation accumulation was found to be significantly slower in genomic sequences directly involved in cellular function, such as exons and gene regulatory sequences. Still, on average, B lymphocytes from aged individuals contained 3-15 damaging mutations in the transcribed part of the exome identified by RNA-seq, plus 15-50 mutations in transcription factor binding sites identified by ATAC-seq. Analysis of the spontaneous mutation spectra in these normal cells revealed signatures similar to those previously found associated with B cell leukemia and other cancers. These results indicate that cancer risk is already part of age-related mutation accumulation in normal cells. Taken together, our single-cell sequencing method provides a firm foundation for analyzing cellular genetic heterogeneity in normal human tissues. Citation Format: Xiao Dong, Lei Zhang, Moonsook Lee, Alexander Y. Maslov, Jan Vijg. Single-cell whole-genome sequencing reveals somatic mutation signatures in normal somatic cells predictive of cancer later in life [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 440.