To identify the core biosynthetic steps of sesquiterpene formation in S. spicatum, a transcriptome established by 454-sequencing of RNA from xylem tissue containing ray parenchyma cells, where sandalwood oil is thought to be synthesised, was explored (Jones et al., 2008). The transcriptome library of 489,364 reads was produced from a single farnesol (8)-rich tree and assembled into 12,537 apparently unique contig sequences. Contigs were classified into functional ontology groups (Fig. S3). More than half (55%) of the contigs had matches of known functions in other species. Of these, the majority (28%) were annotated with ‘‘cellular’’ or ‘‘metabolic processes’’. Genes involved in secondary metabolism were found in ‘‘metabolic processes’’ and ‘‘response to stimuli’’ groups, which comprised 14% and 7% of the transcriptome, respectively. The transcriptome included candidate TPS, P450 and allylic phosphatase sequences, as well as sequences for genes of the MEV and MEP pathways (Table 1). All steps of the MEV pathway, except for phosphomevalonate kinase, were found in the heartwood xylem transcriptome. In contrast, only two enzymes of the MEP pathway were represented. These results are consistent with the MEV pathway being the primary route by which sesquiterpene precursors are produced in plants and the heartwood xylem tissue being particularly rich in sesquiterpenoids. The most abundant transcript of the terpenoid pathway in the xylem transcriptome was HMG-CoA reductase 1 (HMGR1) classified based on sequence relatedness with Arabidopsis thaliana HMGR 1 and HMGR2 (GenBank accession No. AEE35849 and AEC06618, respectively), a critical step for isoprenoid biosynthesis in plants (Chye et al., 1992; Goldstein and Brown, 1990). The MEP pathway, which provides the isoprenoid building blocks for monoterpenes and diterpenes was underrepresented in the transcriptome of S. spicatum heartwood matching the low abundance or lack of these compounds.