Tuberculosis is one of the leading infectious diseases in humans. Discovering new treatments for this disease is urgently required, especially in view of the emergence of multiple drug resistant organisms and to reduce the total duration of current treatments. The synthesis of isoprenoids in Mycobacterium tuberculosis has been reported as an interesting pathway to target, and particular attention has been focused on the methylerythritol phosphate (MEP) pathway comprising the early steps of isoprenoid biosynthesis. In this context we have studied the enzyme 2C-methyl-D-erythritol-4-phosphate cytidylyltransferase (CMS), the third enzyme in the MEP pathway, since the lack of a resolved structure of this protein in M. tuberculosis has seriously limited its use as a drug target. We performed homology modeling of M. tuberculosis CMS in order to provide a reliable model for use in structure-based drug design. After evaluating the quality of the model, we performed a thorough study of the catalytic site and the dimerization interface of the model, which suggested the most important sites (conserved and non-conserved) that could be useful for drug discovery and mutagenesis studies. We found that the metal coordination of CDP-methylerythritol in M. tuberculosis CMS differs substantially with respect to the Escherichia coli variant, consistent with the fact that the former is able to utilize several metal ions for catalysis. Moreover, we propose that electrostatic interactions could explain the higher affinity of the MEP substrate compared with the cytosine 5'-triphosphate substrate in the M. tuberculosis enzyme as reported previously.