Microarray data typically have small numbers of observations per gene, which can result in low power for statistical tests. We investigate permutation based methods for estimating p-values. Data from Affymetrix MOE430v2 microarrays were used to study p-value behavior. Three RMA normalized gene expression datasets were used. Each dataset compared 2 conditions and each condition had 3 biological replicates. Biological detail such as the conditions and the type of biological replicate were not specified.