The gastropod Patella vulgata is abundant on rocky shores in Northern Europe and a significant grazer of intertidal algae. Here we report the application of Illumina sequencing to develop a transcriptome from the adult mantle tissue of P. vulgata. We obtained 47,237,104 paired-end reads of 51 bp, trialled de novo assembly methods and settled on the additive multiple K method followed by redundancy removal as resulting in the most comprehensive assembly. This yielded 29,489 contigs of at least 500 bp in length. We then used three methods to search for candidate genes relevant to biomineralisation: searches via BLAST and Hidden Markov Models for homologues of biomineralising genes from other molluscs, searches for predicted proteins containing tandem repeats and searches for secreted proteins that lacked a transmembrane domain. From the results of these searches we selected 15 contigs for verification by RT-PCR, of which 14 were successfully amplified and cloned. These included homologues of Pif-177/BSMP, Perlustrin, SPARC, AP24, Follistatin-like and Carbonic anhydrase, as well as three containing extensive G-X-Y repeats as found in nacrein. We selected two for further verification by in situ hybridisation, demonstrating expression in the larval shell field. We conclude that de novo assembly of Illumina data offers a cheap and rapid route to a predicted transcriptome that can be used as a resource for further biological study.