Direct immunofluorescence (DIF) is widely used for the diagnosis of bullous diseases and other autoimmune pathologies such as oral lichen planus. There is no evidence in the literature on how the following variants influence the detection rate of DIF: intraoral site chosen for the biopsy, perilesional locus or distant site from the clinical lesion, number of biopsies and instrument used. Objectives: to determine if the following variants influenced the sensitivity (detection rate): intraoral site chosen for the biopsy, perilesional or distant site from the clinical lesion, number of biopsies and instrument used (punch or scalpel). Material and methods: A retrospective study was done at the Cátedra de Patología y Clínica Bucodental II at the Facultad de Odontología, Universidad de Buenos Aires; 136 clinical medical histories were revised for the period March 2000 ? March 2005 corresponding to patients with clinical diagnosis of OLP and bullous diseases (vulgar pemphigus, bullous pemphigoid and cicatricial pemphigoid). Results: DIF detection rate was 65.8% in patients with OLP, 66.7% in cicatricial pemphigoid patients, in bullous pemphigoid 55.6%, in pemphigus vulgaris 100%, and in those cases in which certain diagnosis could not be obtained, the DIF positivity rate was 45.5% (Pearson chi2 (4)= 21.5398 Pr= 0.000). There was no statistically significant difference between the different sites of biopsy (Fisher exact test: 0.825). DIF detection rate in perilesional biopsies was 66.1% and in those distant from the site of clinical lesion was 64.7% (Pearson chi2 (1)= 0.0073 Pr= 0.932). When the number of biopsies were incremented, DIF detection rate also incremented (Pearson chi2 = 8.7247 Pr= 0.003). The biopsies taken with punch had a higher detection rate than those taken with scalpel (39.1% versus 71.7%) (Pearson chi2 = 49.0522 Pr= 0.000). Conclusion: While not statistically significant, the tendency outlined in this study indicates there are intraoral regions in which the detection rate of the DIF technique is higher than others: mouth floor, hard palate, superior labial mucosa, ventral face of tongue. This finding could allow a choice of accessible locations and easy operator manipulation, even in distant places from the clinical lesion. Perilesional biopsies have a detection rate similar to those taken distant from the clinical lesion, and those taken with punch have a higher sensitivity rate than those taken with scalpel (both differences were statistically significant).