Hyper-resistance after a central nervous system injury has been largely referred to as spasticity, which is but one of its neural components. Assessment largely relies on clinical scales (Modified Ashworth scale -MAS and Modified Tardieu scale, MTS) which are unable to distinguish between the non-neural (tissue-related) and the neural (central nervous system-related) components. This study assessed criterion validity and reliability (reproducibility) of muscle stiffness measures, namely, maximum elastic stiffness (ELmax), viscous stiffness (VI), and path length (L-path) in the hand flexor muscles among people with stroke. Measurements were obtained with a wrist-electromechanical oscillatory device (w-EOD). Twenty-four people with arm impairment after stroke were evaluated with the w-EOD and clinical assessment (MAS and MTS), twice on the same day (short-term reliability) and once 10 days later (long-term reliability). For criterion validity, a Spearman coefficient (r) was calculated between stiffness values and the clinical scales. For reliability, intraclass correlation coefficients (ICCs), SEM, and MDC95 were calculated. Moderate correlations were observed between EL max and MAS (r = 0.49) and MTS (V2, r = 0.43; V3, r = 0.49) of the wrist flexors, and finger flexors (MAS, r = 0.60; MTS V2, r = 0.56; MTS V3, r = 0.55). There was a poor correlation between the clinical scales and VI and L-path. Reliability was excellent for all stiffness measurements at short term (EL max : 0.95, VI: 0.94, L-path: 0.92) and good at long term (EL max : 0.87, VI: 0.76, L-path: 0.82). In conclusion, stiffness measurements are valid and reliable to evaluate hyper-resistance in people with stroke.