Objective: This study aimed to evaluate the accuracy and reliability of periapical radiographs obtained with two different projections in determining the cleanliness of root canals in endodontic retreatment using the sectioned roots as a gold standard. Methods:The amount of residual root-filling material after endodontic retreatment procedures in 42 roots was assessed in radiographs and in microscopic photographs of the sectioned roots by five observers. Furthermore, the cleanliness of 80 roots was assessed based on orthogonal and mesio-angulated radiographs. Four parameters were used for the evaluation of cleanliness. Wilcoxon signed-rank tests and percentage agreement calculations were used for comparisons between the radiographs and the gold-standard observations and between the orthogonal and mesio-angulated radiographs, respectively. Intra-and interobserver reproducibility was tested using Kappa statistics and intra-class correlation tests. The statistically significance level was 0.05 Results: In general, significantly more residual root-filling material was detected in the photographs of the split roots (gold standard) than in the radiographs. Comparing orthogonal and eccentric projections, only slight differences in cleanliness were found. The intraobserver reproducibility was fair to almost perfect for radiographs and microscopic photographs. For all parameters, the reproducibility was better for the radiographs than for the microscopic photographs. The interobserver variability ranged from fair to almost perfect agreement for the radiographs. Conclusion: The accuracy of periapical radiographs was poor in determining the cleanliness of root canals after endodontic retreatment, whereas the reliability of the radiographs was fair. A mesio-angulated projection did not contribute essentially to the detection of residual root-filling materials.