ObjectiveTo compare the accuracy of periodontal examination at 6‐sites full‐mouth protocol with clinical and radiographic partial‐mouth recording protocols (PRPs) to diagnose periodontitis considering new 2018 case definition.MethodsPeriodontitis was diagnosed by full‐mouth assessment of clinical attachment level (CAL) at 6‐sites (n = 8529) according to 2018 case definition (standard reference). Three clinical and radiographic PRPs were considered: (i) 4 interproximal sites of all teeth (4‐sites, full‐mouth); (ii) 6‐sites on Community Periodontal Index of Treatment Needs (CPITN) teeth (6‐sites, CPITN teeth); (iii) 4‐sites on CPITN teeth (4‐sites, CPITN teeth); (iv) radiographic examination on interproximal sites of all teeth (rM‐D, full‐mouth); and (v) radiographic examination on interproximal sites of CPITN teeth (rM‐D, CPITN teeth) using Image J®. Sensitivity, specificity, Positive Predictive Value (PPV) and Negative Predictive Value (NPV), accuracy, and Κappa (k) were estimated (α = 0.05).ResultsThe 4‐sites, full‐mouth protocol showed 100% sensitivity and NPV, 79.87% PPV, low specificity (38.32%), 69.30% accuracy, and poor agreement (k = 0.468). 6‐sites and 4‐sites CPITN teeth protocols showed 100% PPV and specificity, sensitivity and NPV of >60%, accuracy of >80%, and good agreement (k > 0.65). All clinical PRPs outperformed radiographic protocols.ConclusionClinical PRPs using index teeth are highly accurate in diagnosing periodontitis considering 2018 case definition.