While current microarray-based hybridization platforms offer a relatively straightforward approach to transcriptome analysis, they are by no means perfect in their design, in part because they provide relative but not absolute quantification of per cell transcript levels and because they require genome or cDNA sequence data for array design. Serial analysis of gene expression (SAGE), a sequencing-based technology, can provide absolute transcript abundances without the a priori need for genome sequence data. However, its general use has been slowed by limited access to high-throughput DNA sequencing centers, the relatively high cost of the sequencing-based analysis, and the slow, labor-intensive set of chemical and enzymatic steps required to prepare a sample for sequencing. I will describe our recent efforts to develop a novel universal microarray platform for SAGE-like analysis of transcript abundances per cell. The technology is universal in that a single array can be used to interrogate any organism or tissue sample of interest. The array is therefore combinatorial in design and displays complementary probes to all possible 9-mer DNA sequences. In general, short-probe microarrays have shown limited success due to the inherent large differences in melting temperatures of AT-rich and GC-rich duplexes. This problem is overcome through the use of locked nucleic acids (LNAs), bicyclic RNA derivatives modified to contain a methylene bridge between the 2'-oxygen and 4'-carbon of the ribose ring. LNA bases exhibit standard Watson-Crick base pairing and can therefore be inserted into any position within a synthetic DNA or RNA oligonucleotide. A group-contribution type molecular thermodynamic model has been developed and used to define the number and position of LNA bases that must be added to each anchored probe to increase the stability of complementary oligonucleotide duplexes on the array to a common melting temperature of 65°C. As LNA containing probes are intolerant of base-pair mismatches, hybridization excludes false positives due to labeled transcripts captured at non-complementary registers. Absolute quantification of transcript abundances is facilitated by precise stoichiometric labeling of each short sequence tag with lanthanide-ion chelates that exhibit very narrow, non-polarized, strongly Stokes-shifted emission bands with extremely long luminescence lifetimes. The performance of our universal microarray platform in providing absolute transcript abundances is demonstrated through direct comparison with SAGE and qPCR data for budding yeast in arrested S-phase of mitosis.