August 2011 saw the release of the first version of the Atlantic cod (Gadus morhua) genome. The assemblies presented were exclusively based on GS FLX Titanium sequencing reads, supplemented with Sanger sequenced BAC ends. More than 22 000 genes were annotated, despite the fragmented genome, by supplementing automatic annotation methods with projections from Stickleback. Sequencing the cod genome is challenging due to the high level of heterozygosity (no inbred specimen is available), and the many short tandem repeats. Here we present the status of the second phase of the cod genome project. This project has as goal to provide the community with an improved genome assembly, allowing for building the gene models directly on the contigs. Longer scaffolds, anchored to a new linkage map, will give pseudochromosomes with a long-range contiguity not available for the published genome. These goals will be achieved by adding high-coverage sequencing data from the Illumina platform, using improved and new assembly and gap-filling algorithms, and utilizing a new, 12k SNP chip to obtain a high-density linkage map. Finally, after automatic annotation, we intend to organise an annotation workshop where we will invite the community for expert curation of the cod genome gene models. We expect the new Atlantic cod genome and annotation to be an invaluable resource for functional and comparative genomics.