Background: Floral scent is one of the important strategies for ensuring fertilization and for determining seed or fruit set. Research on plant scents has hampered mainly by the invisibility of this character, its dynamic nature, and complex mixtures of components that are present in very small quantities. Most progress in scent research, as in other areas of plant biology, has come from the use of molecular and biochemical techniques. Although volatile components have been identified in several orchid species, the biosynthetic pathways of orchid flower fragrance are far from understood. We investigated how flower fragrance was generated in certainPhalaenopsisorchids by determining the chemical components of the floral scent, identifying floral expressed-sequence-tags (ESTs), and deducing the pathways of floral scent biosynthesis inPhalaneopsis bellinaby bioinformatics analysis.Results: The main chemical components in theP. bellinaflower were shown by gas chromatography-mass spectrometry to be monoterpenoids, benzenoids and phenylpropanoids. The set of floral scent producing enzymes in the biosynthetic pathway from glyceraldehyde-3-phosphate (G3P) to geraniol and linalool were recognized through data mining of theP. bellinafloral EST database (dbEST). Transcripts preferentially expressed inP. bellinawere distinguished by comparing the scent floral dbEST to that of a scentless species,P. equestris, and included those encoding lipoxygenase, epimerase, diacylglycerol kinase and geranyl diphosphate synthase. In addition, EST filtering results showed that transcripts encoding signal transduction and Myb transcription factors and methyltransferase, in addition to those for scent biosynthesis, were detected byin silicohybridization of theP. bellinaunigene database against those of the scentless species, rice andArabidopsis. Altogether, we pinpointed 66% of the biosynthetic steps from G3P to geraniol, linalool and their derivatives.Conclusion: This systems biology program combined chemical analysis, genomics and bioinformatics to elucidate the scent biosynthesis pathway and identify the relevant genes. It integrates the forward and reverse genetic approaches to knowledge discovery by which researchers can study non-model plants.