46 punch biopsies of human cartilage were collected and scanned in time-resolved C-scan mode using a scanning acoustic microscope with a single-element 40-MHz transducer. From these data, spectra were estimated from short time gates at regions with low inclination and averaged with respect to their time delay to the tissue surface to account for the curved shape of articular cartilage. The depth-dependent spectral slope and amplitude profiles were used to estimate the maximum values and the depth-dependent slope. The surface reflection was excluded from analysis. In addition, the homodyned k distribution was used to model the signal amplitude of the envelope at tissue depths between 200 and 600 μm. The k and μ parameters were estimated from the envelope data. Kruskal-Wallis analyses of variance were carried out for the acoustic parameters with respect to subscores of the Mankin grading system that are related to cellular and structural abnormalities of the cartilage matrix. The grading was performed on 2D histological cross sections. Additionally the cell number density was assessed.