De Wet, Pieter DaniëlRidel, Alison FanyHagg, Alieske ChristieneHoffman, Jakobus W.L'Abbe, Ericka Noelle2026-04-162026-04-162026De Wet, P.D., Ridel, A.F., Hagg, A.C., Hoffman, J.W. & L'Abbé, E.N. Assessing the reliability of saw mark features: A comparative study using stereomicroscopy, micro-XCT, and 3D printing. Journal of Forensic Sciences 2026; 00: 1–22. https://doi.org/10.1111/1556-4029.70308.10.1111/1556-4029.70308http://hdl.handle.net/2263/109616Presented at the 12th Conference of the African Society of Forensic Medicine (ASFM), Council for Scientific and Industrial Research (CSIR), March 10–13, 2025, in Pretoria, South Africa. DATA AVAILABILITY STATEMENT : The data that support the findings of this study are available from the corresponding author upon reasonable request.Saw mark analysis is an important component of forensic anthropology, particularly in cases of postmortem dismemberment. While stereomicroscopy remains the benchmark for evaluating kerf features, recent advances in imaging technologies offer new opportunities for documenting and interpreting bone trauma. Although these modalities have been studied individually, few investigations have directly compared their relative strengths and limitations. Four fresh human femora were sectioned using six teeth per inch ripsaw to produce 32 false starts and 64 complete cuts. Sixteen kerf features were assessed across stereomicroscopy, micro-focus X-ray computed tomography (micro-XCT), and three-dimensional printed bone models, with scoring for presence, absence, or type, and quantitative measurement where applicable. In addition, a subset of 10 false starts and 10 complete cuts were selected for observer error testing. Minimum kerf width and kerf wall shape showed the highest agreement across observers and modalities, while features such as kerf flare, pull-out striae, and residual kerf wall striae demonstrated lower agreement. Stereomicroscopy provided the highest resolution for fine surface features, whereas micro-XCT enhanced non-destructive visualization of internal kerf morphology and fracture-related traits. Three-dimensional printed models reliably reproduced broader qualitative features but lacked sufficient resolution for fine-scale traits and precise measurements. Significant differences between modalities were observed for several features, emphasizing the influence of visualization method on feature detection and measurement. These findings highlight the need for standardized feature definitions and modality-specific scoring protocols. A multimodal approach of the three modalities is recommended to strengthen the reliability, interpretive value, and forensic applicability of saw mark analysis. HIGHLIGHTS • Observer error evaluation and feature comparison using a single saw across three modalities. • Minimum kerf width, kerf floor, and kerf wall shapes are most reproducible; micro-XCT shows the best agreement. • Stereomicroscopy provided the highest resolution for fine kerf features and striae. • Non-destructive imaging improved detection of spalling and internal kerf traits. • Bone models 3D printed lacked fine detail for analysis but can be used for teaching purposes.en© 2026 The Author(s). Journal of Forensic Sciences published by Wiley Periodicals LLC on behalf of American Academy of Forensic Sciences. This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License.3D printingForensic anthropologyMicro-focus X-ray computed tomography (micro-XCT)Observer errorSaw mark analysisStereomicroscope analysisArticle