Transcriptional pro ling in tissue engineering studies that compare in vitro or in vivo 3D to in vitro 2D samples can be subdivided into three categories. Studies belonging to the rst category tend to focus on a speci c function. For example, Olsavsky et al. (2007) cultured primary hepatocytes and hepatome-derived cell lines HepG2 and Huh7 in a 2D sandwich system, with collagen I as the substratum together with a dilute extracellular matrix (Matrigel™) overlay in a de ned serum-free medium supplemented with dexamethasone at the nanomolar level. Although the study utilized the Affymetrix Human Genome U133Plus 2.0 array, only liver-speci c function gene expressions were reported. Studies belonging to the second category tend to focus on pro ling transcriptions only from cells cultured in 3D scaffolds without a 2D control. For example, Gelain et al. (2006) compared the expression of adult mouse neural stem cells cultured in Matrigel to those cultured in novel peptide scaffolds. The objective was to validate the new scaffolds as opposed to comparing 2D and 3D cultures. In a second example, Ikebe et al. (2007) examined the gene expression pro le of an immortalized human keratinocyte, HaCaT, cultured in a neutralized type I collagen gel, with and without exposure to air for 24 h. The objective was to better understand why an air-liquid interface culture initiated the emulation of in vivo tissue, and as such, a 2D control was not included in the globe gene expression study. Keratinocytes make strati ed epidermoid structures when cultured at an air-liquid interface and have been successfully used in 3D culture studies for more than 25 years (Boyce and Hansbrough, 1988). Studies belonging to the third category not only focused on global gene expression pro les, but also compared expressions in 2D to those in 3D formats. A summary of these studies is compiled in Table 3.1. Although the objectives are all different, a common theme seems to be better understanding the underlying reasons for morphological and functional differences between the two cultures. The purpose of this chapter is to generate insights into the same question with all the studies combined. It was possible to obtain raw data only from Li et al. (2002), Boess et al. (2003), Ghosh et al. (2005), Kenny et al. (2007), and Myers et al. (2008).