DNA microarrays currently provide measurements of sufficiently high quality to allow a wide variety of sound inferences about gene regulation and the coordination of cellular processes to be drawn. Nonetheless, a desire for greater precision in the measurements continues to drive the microarray research community to seek higher measurement quality through improvements in array fabrication and sample labeling and hybridization. We prepared oligonucleotide microarrays by printing 65-mer on aldehyde functional group-derivatized slides as described in a previous study. We could improve the reliability of data by removing enzymatic bias during probe labeling and hybridizing under a more stringent condition. This optimized method was used to profile gene expression patterns for nine different mouse tissues and organs, and multidimensional scaling (MDS) analysis of data showed both strong similarity between like samples and a clear, highly reproducible separation between different tissue samples. Three other microarrays were fabricated on commercial substrates and hybridized following the manufacturer's instructions. The data were then compared with in-house microarray data and reverse transcription-polymerase chain reaction (RT-PCR) data. The microarray printed on the custom aldehyde slide was superior to microarrays printed on commercially available substrate slides in terms of signal intensities, background, and hybridization characteristics. The data from the custom substrate microarray generally showed good agreement in quantitative changes up to 100-fold changes of transcript abundance with RT-PCR data. However, more accurate comparisons will be made as more genomic sequence information is gathered in the public data domain.