By providing genome-scale information on gene expression, microarray technology has gained popularity in diverse areas including clinical medicine. However, the analysis and interpretation of microarray data are often complicated. This chapter describes various strategies for microarray data analysis. The analysis starts with the scanned image of a microarray. The image information is processed and summarized to numerical values that represent the abundance of transcripts. Technical variability and systematic biases can be minimized with the proper procedures of background correction and normalization. Considerable numbers of genes are not expressed or not detected by microarray technology. Those genes can be filtered out before further statistical comparison to reduce the dimensionality of the problem. The next step in analysis involves statistical comparison, cluster analysis, and visualization. Genes from the same cluster are considered to be coexpressed and/or coregulated. Also, we can group coexpressed genes into categories by their biological function and cellular location. By combining prior knowledge and statistical results, we can make an inference based on the gene expression profiles.