This study aimed to validate finite element analysis (FEA) estimation of strength, identify high-resolution peripheral quantitative computed tomography (HR-pQCT) measures correlating with strength, and evaluate the precision of HR-pQCT measurements to longitudinally monitor effects of osteoporosis treatment in ovariectomized (OVX) non-human primates (NHPs). HR-pQCT images were acquired in three groups of NHPs: Intact (n = 10), OVX-odanacatib treated (OVX-ODN 30 mg/kg, n = 10) and OVX-vehicle treated (OVX-Veh, n = 10) at the ultradistal (UD) and distal 1/3 radii and tibia at 12, 16 and 20 months. FEA estimates of bone strength using the Pistoia criterion were validated by ex-vivo mechanical compression (r2 = 0.95) of the UD radius. Single linear regressions of FEA-determined ultimate stress showed high correlation with HR-pQCT derived parameters: integral vBMD (r2 = 0.86), bone volume fraction (r2 = 0.84) and cortical thickness (r2 = 0.79). Precision of HR-pQCT measurements, obtained from an excised radius and tibia, showed low variation (CV = 0.005%–5.6%) and helped identify possible sources of error. Comparison of OVX-Veh and Intact groups showed decreases in bone parameters demonstrating trends consistent with bone loss. Comparison of OVX-ODN and OVX-Veh groups showed a treatment effect with increases in bone parameters: integral vBMD (477 ± 27 vs. 364 ± 22 mg HA/cm3) and cortical thickness (Ct.Th) (0.90 ± 0.07 vs. 0.64 ± 0.04 mm) at the UD radius, Ct.Th (2.15 ± 0.28 vs. 1.56 ± 0.08 mm) at the distal 1/3 radius. Axial compression peak stress calculated and obtained experimentally showed the OVX-ODN group was 33% stronger than the OVX-Veh group. We conclude that HR-pQCT and FEA serve as robust techniques to longitudinally monitor bone parameters and strength in NHP's.