Purpose: This study aimed to assess the interdental bone level in premolar bitewing radiographs while retracting the cheeks. Methods: Seventy-two horizontal bone defects were created on dried mandibles and maxillae. The distance from the bone level to the cement-enamel junction of premolars was detected by a modified digital caliper (considered the gold standard). The reliability of all radiographs was assessed by intraclass correlation coefficient (ICC), and the validity was compared to the gold standard using the analysis of variance test. P-values less than 0.05 were considered statistically significant. Results: This study showed that the reliability of radiographs without a cheek simulator and with 0.16 second exposure time was significantly higher than that of the two other groups (ICC=0.96 compared to 0.93 and 0.88, respectively). The results from the radiographs without a cheek simulator and with 0.16 second exposure time were more similar to the gold standard measures than those of the two other groups, although the difference was not statistically significant. Conclusions: Retracting the buccal soft tissue plays an important role in increasing the accuracy of radiographs in detecting the interdental alveolar bone level and produces more accurate results than increasing the exposure time, although it does not have a significant role in reliability of results.