Efficacy of custom-made foot orthoses (CFO) largely depends on geometric characterization and outcomes can be suboptimal due to insufficient quantitative insights thereof. Addressing this is crucial for understanding geometric characteristics in dosage-response modelling. This study evaluates the reliability of a handheld 3D scanning system for foot digitization, focusing on traditional and advanced geometric features. Two examiners used a handheld 3D scanner to capture the foot shape of thirty healthy participants in a non-weightbearing condition. Intra-tester, test-retest and inter-tester reliabilities of multiple zero-dimensional (0D) anthropometric and geometric variables were evaluated. Subsequently, the consistency of capturing the geometry of the plantar foot surface was assessed via quantification of one-dimensional (1D) anteroposterior and mediolateral plantar surface curvatures. Intraclasss correlation coefficients of the 0D parameters showed good to excellent intra- and inter-tester reliability. A pairwise comparison of the 1D curvatures showed good consistency on intra-tester level. These results demonstrated that 0D anthropometric measurements of the human foot, obtained using a handheld 3D scanning system in a non-weightbearing condition, exhibit good to excellent reliability. Additionally, the 1D curves related to the plantar surface showed high consistency, supporting its usage to quantify plantar geometry and provide insights into the role of geometric characteristics in dosage-response modelling in CFO practice.