This study aimed to evaluate the in vitro performance in detecting caries around composite restorations in permanent posterior teeth and the impact of treatment decisions of two visual criteria: International Dental Federation (FDI) criteria and the Caries Associated with Restorations and Sealants (CARS) system. The correlation among secondary caries and the presence of gap measured with a three-dimensional (3D) intraoral scanner was also aimed. One hundred sixteen teeth were assessed by a trained and calibrated examiner according to the FDI criteria or CARS system. A second examiner measured the gap on the 3D models using specific software. The reference standard was the histological examination performed by a third examiner blind to the other evaluations. Other 30 extracted permanent teeth were used only to assess the reproducibility of the methods. The same random sample was selected for re-examination by all three methods, and there were seven days between the examinations. Unweighted and weighted kappa tests were conducted to assess intra-examiner reproducibility. Spearman's correlation coefficient (Rho) and 95% confidence intervals (95%CI) were calculated between the histological examination and scores obtained with FDI criteria, CARS system and treatment decisions. Spearman correlation between the visual and scanner evaluation with the reference standard was performed. Spearman's rank correlation analyses were conducted independently between the gap evaluated and measured by the visual inspection with the gap assessed using the scanner. The reproducibility of the visual score systems reached high values. Spearman's correlation coefficients (Rho; 95%CI) between the following variables versus histology were: FDI presence of caries (0.65; 0.53 to 0.74); CARS scores (0.65; 0.52 to 0.74); FDI treatment decision (0.46; 0.31 to 0.59); and CARS treatment decision (0.62; 0.49 to 0.72). Rho (95%CI) between histology and the gap assessment by the visual inspection was 0.59 (0.45 to 0.70), with the gap measurement by the visual inspection was 0.49 (0.33 to 0.62), and the gap measured by the scanner was 0.37 (0.18 to 0.53). Both visual criteria present similar performance in detecting caries. The correlation among treatment decisions is moderate for the FDI and CARS criteria, and both are moderately correlated with lesion depth, with a slight CARS superiority. However, visual examination presents better performance than the 3D intraoral scanner on gap size assessment