Genome wide gene expression analysis by cDNA microarrays is often limited by minute amounts of starting RNA. We therefore tested an optimized linear RNA amplification protocol using the RiboAmp<sup>®</sup> amplification kit in the setting of cDNA microarrays. We isolated mRNA from a human kidney cell line (HK-2; ATCC) and from Universal Human Reference RNA (STR; Stratagene). After performing one and two rounds of linear RNA amplification, respectively, the amplified RNAs were co-hybridized to cDNA microarrays. Linearity and reproducibility of the individual experiments were then assessed by calculating the Pearson correlation. The intra-amplification consistency showed a correlation of 0.968 for the first round, 0.907 for the second round and 0.912 for two successive rounds of amplification. If the first round was compared to unamplified material, r was 0.925. The second round amplification yielded a correlation of 0.897 if compared to unamplified mRNA. Two rounds of amplification starting from 200 pg of mRNA compared to unamplified material resulted in a correlation of 0.868. These results indicate that linear amplification using RiboAmp<sup>®</sup> kit yields amplified RNA with a high degree of linearity and reproducibility.