A number of non-tuberculous mycobacterium species are opportunistic pathogens and ubiquitously form biofilms. These infections are often recalcitrant to treatment and require therapy with multiple drugs for long duration. The biofilm resident bacteria also display phenotypic drug tolerance and thus it has been hypothesized that the drug unresponsiveness in vivo could be due to formation of biofilms inside the host. We have discussed the biofilms of several pathogenic non-tuberculous mycobacterium (NTM) species in context to the in vivo pathologies. Besides pathogenic NTMs, Mycobacterium smegmatis is often used as a model organism for understanding mycobacterial physiology and has been studied extensively for understanding the mycobacterial biofilms. A number of components of the mycobacterial cell wall such as glycopeptidolipids, short chain mycolic acids, monomeromycolyl diacylglycerol, etc. have been shown to play an important role in formation of pellicle biofilms. It shall be noted that these components impart a hydrophobic character to the mycobacterial cell surface that facilitates cell to cell interaction. However, these components are not necessarily the constituents of the extracellular matrix of mycobacterial biofilms. In the end, we have described the biofilms of Mycobacterium tuberculosis (Mtb), the causative agent of tuberculosis. Three models of Mtb biofilm formation have been proposed to study the factors regulating biofilm formation, the physiology of the resident bacteria, and the nature of the biomaterial that holds these bacterial masses together. These models include pellicle biofilms formed at the liquid-air interface of cultures, leukocyte lysate-induced biofilms, and thiol reductive stress-induced biofilms. All the three models offer their own advantages in the study of Mtb biofilms. Interestingly, lipids (mainly keto-mycolic acids) are proposed to be the primary component of extracellular polymeric substance (EPS) in the pellicle biofilm, whereas the leukocyte lysate-induced and thiol reductive stress-induced biofilms possess polysaccharides as the primary component of EPS. Both models also contain extracellular DNA in the EPS. Interestingly, thiol reductive stress-induced Mtb biofilms are held together by cellulose and yet unidentified structural proteins. We believe that a better understanding of the EPS of Mtb biofilms and the physiology of the resident bacteria will facilitate the development of shorter regimen for TB treatment.