Microarray analysis has provided a novel means of identifying clues into the mechanisms of disease development. As a methodology, microarray analysis holds the promise for genome-wide screening in which 2 tissues (diseased and normal) are compared, and molecular pathways that defined the phenotype of the disease could be precisely defined. Alternatively, microarray experiments can be used to differentially compare pathologically similar diseased tissues to predict response to chemotherapy and risk of recurrence. However, the clinician should be aware that various sources of error can influence microarray analysis results. Sources of error can be minimized but not eliminated, explaining why meticulously conducted experiments in different laboratories or using different platforms result in different lists of genes. Confirmation and validation of genome-wide microarray results using ancillary methods remains a critical step. With proper confirmatory studies and cautious interpretation, microarray analysis represents a powerful tool for molecular discovery.