Here it is presented a pipeline for obtaining a reliable transcriptome in non-model species, such as pine and sole, using NGS reads from natural, heterozygotic populations. The pipeline includes pre-processing, assembly, reconciliation, annotation and post-processing using well-know software (Oases, MIRA3, Euler-SR, CAP3, CD-Hit, ABySS, etc.) as well as new software developed by authors (SeqTrimNext, Full-LengtherNext, Sma3Annot). Reference transcriptomes obtained with this approach have been used for cloning genes, printing and perform microarrays, and carry out RNA-Seq analyses that were confirmed by RT-PCR. These successful applications suggest that the pipeline is adequate non-model organisms.