AIM: To assess the impact of cross-sectional root canal shape (CSRCS) on the canal volume that can be filled and the root filling material that remains following a subsequent retreatment procedure. METHODOLOGY: A total of 15 extracted two-rooted human maxillary premolars and 15 mandibular first molars were used. Both root canals in the premolars (N = 30) and the distal root canal in the molars (N = 15) were prepared using ProFile instruments and filled by lateral compaction using gutta-percha and AH Plus sealer. Canals were later retreated using the last ProFile used for instrumentation followed by two ProFiles of increasing size. Teeth were viewed in a μCT scanner before and after each treatment step. Defined and validated threshold levels were used to differentiate empty root canal volumes, root dentine and root filling materials from each other. CSRCS was defined as the averaged ratio between bucco-lingual and mesio-distal canal diameter (round ≤ 1, oval 1-2, long oval 2-4 and flattened ≥ 4), determined for each 1 mm over the total root length. Data were averaged between the two canals in premolars, only the distal canals were assessed in molars. Parametric and non-parametric tests were used to statistically compare the data, alpha = 0.01. RESULTS: Canals in premolars had a round CSRCS after preparation (1.0 ± 0.0), whereas distal counterparts in molars were oval (1.6 ± 0.5). Significantly (P < 0.01) more canal volume could be filled, and significantly less filling material remained after retreatment in premolars compared with mandibular molar distal canals. There was a high correlation between CSRCS, filled canal volume and remaining filling material. CONCLUSIONS: The endodontic procedures under investigation were significantly influenced by the cross-sectional root canal shape.