This in vitro study aimed to compare the 3-dimensional (3D) accuracy of intraoral and desktop scanners when digitizing the post and core polyvinyl siloxane impressions. Ten extracted human teeth were prepared to build a post space in the root canal. Each tooth was scanned using a micro-computed tomography device (MCT), and the slice data were reconstructed as controls. A conventional quadrant polyvinyl siloxane impression was made and scanned thrice using an intraoral scanner (IOS) and a desktop blue-light scanner (IMS), with one of the scans randomly selected as the definitive cast. Precision was assessed by comparing the deviations among repeated scans. Trueness was evaluated by analyzing the deviations between the MCT and each definitive cast (MCT-IOS and MCT-IMS). The clinically acceptable root mean square (RMS) value was set to 50 µm. The mean RMS values of the repeated scans of IOS and IMS were 18.3 ± 3.9 µm and 13.9 ± 6.2 µm (P < 0.05), and the difference between them was 4.4 ± 5.9 µm. The mean RMS values of MCT-IOS and MCT-IMS were 28.3 ±3.0 µm and 31.3 ± 4.2 µm (P < 0.05), with a difference of 3.0 ± 2.9 µm. The RMS values were significantly lower than 50 µm (P < 0.05). The IOS showed slightly better trueness than the IMS but slightly lower precision. The precision and trueness of the two scanners for digitizing post and core impressions were acceptable for clinical application.