Diagnosing scaphoid fractures remains challenging. High-resolution peripheral quantitative computed tomography (HR-pQCT) might be a potential imaging technique, but no data are available on its feasibility to scan the scaphoid bone in vivo. Patients (≥18 years) with a clinically suspected scaphoid fracture received an HR-pQCT scan of the scaphoid bone (three 10.2-mm stacks, 61-μm voxel size) with their wrist immobilized with a cast. Scan quality assessment and bone contouring were performed using methods originally developed for HR-pQCT scans of radius and tibia. The contouring algorithm was applied on coarse hand-drawn pre-contours of the scaphoid bone, and the resulting contours (AUTO) were manually corrected (sAUTO) when visually deviating from bone margins. Standard morphologic analyses were performed on the AUTO- and sAUTO-contoured bones. Ninety-one patients were scanned. Two out of the first five scans were repeated due to poor scan quality (40%) based on standard quality assessment during scanning, which decreased to three out of the next 86 scans (3.5%) when using an additional thumb cast. Nevertheless, after excluding one scan with an incompletely scanned scaphoid bone, post hoc grading revealed a poor quality in 14.9% of the stacks and 32.9% of the scans in the remaining 85 patients. After excluding two scans with contouring problems due to scan quality, bone indices obtained by AUTO- and sAUTO-contouring were compared in 83 scans. All AUTO-contours were manually corrected, resulting in significant but small differences in densitometric and trabecular indices (<1.0%). In vivo HR-pQCT scanning of the scaphoid bone is feasible in patients with a clinically suspected scaphoid fracture when using a cast with thumb part. The proportion of poor-quality stacks is similar to radius scans, and AUTO-contouring appears appropriate in good- and poor-quality scans . Thus, HR-pQCT may be promising for diagnosis of and microarchitectural evaluations in suspected scaphoid fractures.