A major obstacle to studying the functions of particular gene products in the mouse-tick infectious cycle of Borrelia burgdorferi has been an inability to knock out genes in pathogenic strains. Here, we investigated conditions for site-directed mutagenesis in B31 MI, the low-passage-number, infectious B. burgdorferi strain whose genome was sequenced. We inactivated several plasmid and chromosomal genes in B31 MI and determined that clones carrying these mutations were not infectious for mice. However, we found extensive heterogeneity among clones and mutants derived from B31 MI based on colony phenotype, growth rate, plasmid content, protein profile, and transformability. Significantly, several B31 MI clones that were not subjected to mutagenesis but that lacked particular plasmids also exhibited defects at various stages in the infectious cycle. Therefore, the high degree of clonal polymorphism within B31 MI complicates the assessment of the contributions of individual genes to the observed phenotypes of the mutants. Our results indicate that B31 MI is not an appropriate strain background for genetic studies in infectious B. burgdorferi, and a well-defined isogenic clone is a prerequisite for targeted mutagenesis. To this end, we derived several wild-type clones from B31 MI that were infectious for mice, and gene inactivation was successful in one of these clones. Due to the instability of the genome with in vitro propagation, careful monitoring of plasmid content of derived mutants and complementation of inactivated genes will be crucial components of genetic studies with this pathogen.