Premise of the study: We present a protocol for the annotation of transcriptome sequence data and the identification of candidate genes therein using the example of the nonmodel conifer Abies alba. Methods and Results: A normalized cDNA library was built from an A. alba seedling. The sequencing on a 454 platform yielded more than 1.5 million reads that were de novo assembled into 25 149 contigs. Two complementary approaches were applied to annotate gene fragments that code for (1) well-known proteins and (2) proteins that are potentially adaptively relevant. Primer development and testing yielded 88 amplicons that could successfully be resequenced from genomic DNA. Conclusions: The annotation workflow offers an efficient way to identify potential adaptively relevant genes from the large quantity of transcriptome sequence data. The primer set presented should be prioritized for single-nucleotide polymorphism detection in adaptively relevant genes in A. alba.