Microarrays have become one of the leading technologies used for gene expression analysis and functional genomics in many biological fields. Potential applications of microarrays can facilitate advances in molecular biology, systems biology, functional genomics, clinical medicine, and pharmacogenomics. However, microarray data can also lead to inaccurate and irreproducible conclusions. Here, we present a critical review of current computational tools used for normalization, statistical analysis, cluster analysis, and mathematical modeling-based analysis. Despite the pitfalls and challenges that still encompass the computational analysis of microarray data, the use of this technology remains very promising. In our opinion, achieving the full potential of microarray technology requires additional theoretical advances.