The authors describe a large-scale, random approach called reduced representation bisulfite sequencing (RRBS) for analyzing and comparing genomic methylation patterns. They size-selected BglII restriction fragments to 500–600 bp, equipped them with adapters, treated them with bisulfite, PCR amplified, cloned, and sequenced the libraries. RRBS libraries were constructed from murine embryonic stem (ES) cells and from ES cells lacking DNA methyltransferases Dnmt3a and 3b, as well as with knocked-down levels of Dnmt1. Sequencing of 960 RRBS clones from Dnmt[1kd,3a−/−,3b−/−] cells generated 343 kb of non-redundant bisulfite sequence covering 66,212 cytosines in the genome. The results showed that all but 33 cytosines had been converted to uracil, indicating a conversion rate of >99.9%. Of the remaining cytosines, 35 were found in CpG and 3 in CpT dinucleotides. Non-CpG methylation was >250-fold reduced compared with wild-type ES cells, suggesting a role for Dnmt3a and/or Dnmt3b in CpA and CpT methylation. Further analysis revealed no consensus sequence around the methylated sites or evidence of clustered residual methylation in the genome, indicating random loss rather than specific maintenance of methylation in Dnmt[1kd,3a−/−,3b−/−] cells. The near-complete bisulfite conversion and unbiased representation of RRBS libraries suggest that random shotgun bisulfite sequencing can be scaled to a genome-wide approach.The authors describe a large-scale, random approach called reduced representation bisulfite sequencing (RRBS) for analyzing and comparing genomic methylation patterns. They size-selected BglII restriction fragments to 500–600 bp, equipped them with adapters, treated them with bisulfite, PCR amplified, cloned, and sequenced the libraries. RRBS libraries were constructed from murine embryonic stem (ES) cells and from ES cells lacking DNA methyltransferases Dnmt3a and 3b, as well as with knocked-down levels of Dnmt1. Sequencing of 960 RRBS clones from Dnmt[1kd,3a−/−,3b−/−] cells generated 343 kb of non-redundant bisulfite sequence covering 66,212 cytosines in the genome. The results showed that all but 33 cytosines had been converted to uracil, indicating a conversion rate of >99.9%. Of the remaining cytosines, 35 were found in CpG and 3 in CpT dinucleotides. Non-CpG methylation was >250-fold reduced compared with wild-type ES cells, suggesting a role for Dnmt3a and/or Dnmt3b in CpA and CpT methylation. Further analysis revealed no consensus sequence around the methylated sites or evidence of clustered residual methylation in the genome, indicating random loss rather than specific maintenance of methylation in Dnmt[1kd,3a−/−,3b−/−] cells. The near-complete bisulfite conversion and unbiased representation of RRBS libraries suggest that random shotgun bisulfite sequencing can be scaled to a genome-wide approach.