The diversity of criteria used in the diagnosis of periodontitis in pregnant women makes it difficult to define and compare the disease. Using a systematic review, this study evaluated the accuracy of criteria for diagnosing periodontitis in pregnant women. Searches were carried out in the databases: Medline/PubMed, Embase, Scopus, Web of Science, SciELO, Lilacs, ProQuest, and CINAHL. Validation studies of the criteria used for diagnosing periodontitis met the inclusion criteria. The study steps were performed by three independent reviewers. A qualitative synthesis of the included articles was carried out and the quality methodological analysis using the QUADAS-2 instrument. The assessment of the quality of the evidence was obtained through the GRADE tool. A total of 592 records were identified, of which only 6 made up this systematic review. The included studies analyzed different criteria for diagnosing periodontitis, evaluating 3,005 pregnant women. The criteria selected as a test presented results equivalent to the performance of those recognized as the gold standard. The self-reported criteria were of lower diagnostic accuracy. A major limitation of this review was the small number of primary studies that evaluated clinical diagnosis of periodontitis in pregnant women, which was highly heterogeneous, making it impossible to carry out accuracy meta-analysis and additional analyzes. There is a lack of consensus on the criteria for the diagnosis of periodontitis, with great variability in the accuracy and prevalence of the disease in pregnant women.