The recent completion of a draft genome sequence for Methanosarcina barkeri has allowed the application of various high throughput post-genomics technologies, such as nucleic acid microarrays and mass spectrometry of proteins to detect global changes in transcription and translation that occur in response to experimental treatments. However, due to the production of a thick heteropolysaccharide outer layer, M. barkeri usually grows in large aggregates of cells rather than as individual, planktonic cells. Complete disruption of these aggregates and lysis of the released cells presents technical difficulties in ensuring the isolation of intact RNA from the entire population of cells. Initial attempts at isolating RNA from M. barkeri using several standard extraction protocols gave incomplete lysis of cells and resulted in low yields of poor quality RNA. In this study, we tested several chemical and mechanical disruption modifications of standard RNA extraction methods to optimize the extraction efficiency and minimize the number of unlysed cells remaining after extraction. As a further test of the quality of the resulting RNAs, their performance in replicate microarray analyses were determined. The results showed that inclusion of a liquid nitrogen grinding step prior to Trizol extraction, combined with moderate bead beating, yielded the most complete cell lysis, the highest yield of RNA and the most reproducible microarray results for M. barkeri. From these results it is clear that the methods used to isolate RNA can have a significant impact on the variability, trend and, presumably, the accuracy of microarray data. In addition, functional analysis of the microarray results obtained with RNA from the optimized protocol showed that, as expected, the genes involved in methanogenesis were among the most highly expressed genes in M. barkeri.