Microarray technologies allow the simultaneous monitoring of gene transcript abundance at a magnitude that allows whole genome assessment of every gene in an organism. When microarray technologies first arrived on the scientific scene, they seemed like such a great idea and the answer for everything -so much data; so many questions to ask, and so fast! Whether one was exploring data or testing a hypothesis, microarray technologies provided a vast amount of data around which many stories were told; some validated by other scientific means, some not. The probability of finding genes that were differentially expressed (i.e., something caused a change in the gene's transcript abundance between conditions) was almost certain, especially if you were testing every gene in the genome (or, at least hundreds), and usually the results were encouraging. But after awhile, and after a bit of thought, the scientific questions quickly turned to questions about whether the observed changes were actual changes in the transcript abundance or just random noise, when and why were statistical experiment design and analysis methods necessary, and to what extent is independent experimental confirmation necessary. A less obvious question for many biologists is what does statistical significance mean when biological significance is at the heart of the application?