We compared seedling and reproductive stage transcriptomes for both coconut types to determine potential molecular mechanisms underlying control of flowering time in coconut.Approximately 6.6 Gb of unmapped clean reads were generated for each of four samples: tall coconut seedlings, tall coconut mature trees, dwarf coconut seedlings and dwarf coconut mature trees. After mapping all clean reads to the coconut reference genome using the software HISAT, a total of 36 558 transcripts were predicted. Among them, a total of 7 037 transcripts were predicted to be non-coding RNA. The remaining 29 521 were identified as protein-coding transcripts, 25 114 (85%) of which were consistent with previously known proteins predicted based on the coconut genome and the remaining 4407 of which were novel protein-coding transcripts.