Pemphigus is a unique and interesting autoimmune disease, in which autoantibodies play a major pathogenic role and cause blister formation. Several questions raised from clinical observation in pemphigus have been answered with logic at the molecular level owing to recent remarkable progress in research in the field of pemphigus. The clinical phenotype of classic pemphigus, pemphigus vulgaris (PV) and pemphigus foliaceus (PF), is defined by anti-desmoglein autoantibody profile. Sera containing anti-Dsg3 IgG alone cause mucosal dominant PV with limited skin involvement. Sera containing both anti-Dsg3 and anti-Dsg1 cause mucocutaneous PV, which affects both the skin and mucous membrane. Sera containing only anti-Dsg1 cause PF, which shows cutaneous but no mucosal involvement. In herpetiform pemphigus (HP) most sera recognize Dsg1 and the rest of them recognize Dsg3, indicating that HP is a clinical variant of PF or PV. Patients with paraneoplastic pemphigus (PNP) have autoantibodies against multiple molecules. Now we know that they have autoantibodies against all members of the plakin family, which are cytoplasmic proteins and include desmoplakin, BPAG1, envoplakin, periplakin, and plectin. Cell surface target antigens of PNP, which blister-inducing pathogenic autoantibodies attack, were finally discovered to be Dsg3 and Dsg1. Therefore, PNP is characterized as an autoimmune disease against plakin molecules and desmogleins. Autoimmune targets of IgA pemphigus are likely more heterogeneous than originally thought. So far, desmocollin 1, Dsg3, and Dsg1 are known as their target antigens. Thus, pemphigus has become one of well-characterized tissue-specific autoimmune diseases. Pemphigus will be a good model disease in the next century to address the central issue of autoimmune disease and basic immunology; why and how do patients with autoimmune diseases start to recognize self as non-self?