Current methods for the functional analysis of microarray gene expression data make the implicit assumption that genes with similar expression profiles have similar functions. However genes involved in the same biological pathway could be associated because of their temporal and spatial relationship but not always because of the high expression similarity. Here we explored the possible application of shortest-path analysis on expression of rice genes under abscisic-acid, drought, cold and salinity stress by rice cDNA microarray. We constructed the shortest path among the selected genes with low expression similarity using software we developed and found functional relationships among rice genes and predicted the function of unknown genes. The data were explored to understand the network of gene expression and response to environmental stress in rice.