Plant breeding is enhanced by the availability of molecular markers for rapid screening and selection in populations. Identification of polymorphic loci in cultivated tomato (Solanum lycopersicum L.) has been hampered by limited genome sampling across cultivated types. Whole transcriptome sequencing of six accessions that span cultivated market classes was performed using sequencing by synthesis. A total of 291,915,037 quality filtered reads representing 17 Gb of sequence were generated. Assembly of the reads resulted in 30.6 to 34.9 Mb of sequence for each of the six accessions and provided representation of 55.3 to 59.6% of the predicted tomato gene set with a wide range of molecular function Gene Ontologies (GOs) represented. A computational pipeline was developed to identify single nucleotide polymorphisms (SNPs). When coupled with two Sanger‐derived expressed sequence tag datasets and a reference genome, 62,576 nonredundant putative SNPs in tomato were identified. The SNPs within the contigs were present within all of the GO molecular function categories. The computational pipeline had validation rates in SNP genotyping assays that ranged from 95 to 100%, and the utility of these SNPs for assessing genetic variation within cultivated and wild populations was demonstrated. Collectively, the transcript sequences and the annotated SNPs provide a resource to facilitate tomato genetics and breeding efforts.