Background: Paraneoplastic pemphigus (PNP) is an autoimmune blistering disease characterized by the production of autoantibodies mainly directed against proteins of the plakin family. An overlapping distribution of autoantibody specificities has been recently reported between PNP, pemphigus vulgaris (PV), and pemphigus foliaceus (PF), which suggests a relationship between the different types of pemphigus. Objective: Our purpose was to evaluate the sensitivity and the specificity of clinical, histologic, and immunologic features in the diagnosis of PNP. Methods: The clinical, histologic, and immunologic features of 22 PNP patients were retrospectively reviewed and compared with those of 81 PV and PF patients without neoplasia and of 8 PV and 4 PF patients with various neoplasms. Results: One clinical and 2 biologic features had both high sensitivity (82%-86%) and high specificity (83%-100%) whatever the control group considered: (1) association with a lymphoproliferative disorder, (2) indirect immunofluorescence (IIF) labeling of rat bladder, and (3) recognition of the envoplakin and/or periplakin bands in immunoblotting. Two clinicopathologic and two biologic features had high specificity (87%-100%) but poor sensitivity (27%-59%): (1) clinical presentation associating erosive oral lesions with erythema multiforme-like, bullous pemphigoid-like, or lichen planus-like cutaneous lesions; (2) histologic picture of suprabasal acantholysis with keratinocyte necrosis, interface changes, or lichenoid infiltrate; (3) presence of both anti-epithelial cell surface and anti-basement membrane zone antibodies by IIF; and (4) recognition of the desmoplakin I and/or BPAG1 bands in immunoblotting. Interestingly, 45% of patients with PNP presented initially with isolated oral erosions that were undistinguishable from those seen in PV patients, and 27% had histologic findings of only suprabasal acantholysis, which was in accordance with the frequent detection of anti-desmoglein 3 antibodies in PNP sera. Conclusion: The association with a lymphoproliferative disorder, the IIF labeling of rat bladder, and the immunoblotting recognition of envoplakin and/or periplakin are both sensitive and specific features in the diagnosis of PNP. (J Am Acad Dermatol 2000;43:619-26.)