Parasteatoda tepidariorum is an increasingly popular model for the study of spider development and the evolution of development more broadly. However, fully understanding the regulation and evolution of P. tepidariorum development in comparison to other animals requires a genomic perspective. Although research on P. tepidariorum has provided major new insights, gene analysis to date has been limited to candidate gene approaches. Furthermore, the few available EST collections are based on embryonic transcripts, which have not been systematically annotated and are unlikely to contain transcripts specific to post-embryonic stages of development. We therefore generated cDNA from pooled embryos representing all described embryonic stages, as well as post-embryonic stages including nymphs, larvae and adults, and using Illumina HiSeq technology obtained a total of 625,076,514 100-bp paired end reads. We combined these data with 24,360 ESTs available in GenBank, and 1,040,006 reads newly generated from 454 pyrosequencing of a mixed-stage embryo cDNA library. The combined sequence data were assembled using a custom de novo assembly strategy designed to optimize assembly product length, number of predicted transcripts, and proportion of raw reads incorporated into the assembly. The de novo assembly generated 446,427 contigs with an N50 of 1,875 bp. These sequences obtained 62,799 unique BLAST hits against the NCBI non-redundant protein data base, including putative orthologs to 8,917 Drosophila melanogaster genes based on best reciprocal BLAST hit identity compared with the D. melanogaster proteome. Finally, we explored the utility of the transcriptome for RNA-Seq studies, and showed that this resource can be used as a mapping scaffold to detect differential gene expression in different cDNA libraries. This resource will therefore provide a platform for future genomic, gene expression and functional approaches using P. tepidariorum.