The variety of criteria used to establish a clinical diagnosis of periodontal disease reflects the difficulty in selecting a classification of the disease, as well as in comparison of the findings of the scientific studies in periodontics. OBJECTIVES: to compare partial and full-mouth periodontal examinations in the establishment of periodontal disease (PD) diagnosis. METHODS: After a full-mouth periodontal examination (FPE), 214 individuals were classified as periodontal disease subjects according to the following criteria: at least 4 or more teeth, with one or more sites with probing depth ≥ 4 mm, with clinical attachment loss ≥ 3 mm at the same site (Lopez et al., 2002; Gomes-Filho et al., 2005). The same individuals were classified as periodontal disease subjects by means of the partial periodontal examination (PPE), according to the periodontal attachment loss index (CPI/OMS), and the following criterion: at least one of the index teeth (17,16,47,11,31,26,27,36, and37) with one or more sites with clinical attachment loss ≥ 4 mm in relation to the cementum-enamel junction. This first criterion was used as the gold standard to obtain the PPE diagnostic values: sensitivity, specificity, and positive and negative predictive values. RESULTS: The results showed a PD prevalence of 46% to the PPE and of 17% to the FPE. Furthermore, the following PPE diagnostic values were observed in relation to the FPE: sensitivity 100%, 95% CI [88.6, 100]; specificity 64%, 95% CI [57.2, 71%]; positive predictive value 38%, 95% CI [28.6, 48.3], and negative predictive value 100%, 95% CI [95.9, 100]. CONCLUSIONS: The results suggest that the PPE distorts PD prevalence. Furthermore, the criterion choice depends on the study design and the sample size, since the PPE has the capacity to identify an impressive number of false-positives. This situation should be further investigated in studies with small samples as well as in those investigating the hypothesis of association, because more accurate classification measurements are needed.