Inflammatory bowel diseases are becoming increasingly common throughout the world, both in developed countries and increasingly in rapidly developing countries. Multiple lines of evidence point to a role for the microbial composition of the gastrointestinal tract in the etiology of IBD, but to date, attempts to define a specific microbial cause for IBD have proved unsuccessful. Microbial 16S rRNA profiling shows that IBD patients have elevated levels of Enterobacteriaceae, in particular Escherichia coli, and reduced levels of Faecalibacterium prausnitzii. The observed E. coli have been assigned to a specific pathovar, adherent-invasive E. coli (AIEC). Adherent-invasive E. coli are a genomically heterogenous group, and whereas many groups have attempted to identify specific genetic markers that differentiate AIEC from non-AIEC strains, very few concrete genetic associations have been uncovered. Here, we highlight the advantages of applying a phenotyping approach to the study of these organisms, rather than solely depending on a sequencing or genomic-based screening strategy because virulence-associated phenotypes exhibit behaviors of emergent systems. In this respect, attempts at genetic reductionism are prone to failure because there are numerous metabolic, regulatory or genetic paths that can underlie these virulence-associated behaviors. Here, we review these IBD-associated phenotypes in E. coli and make recommendations for experimental approaches to advance our understanding of IBD-associated bacteria more generally. With advances in high-throughput screening and nongenetically based metabolomic characterization of IBD-associated bacteria, we anticipate a fuller understanding of how altered microbial communities contribute to the development of IBD.