Structural alterations associated with early, mostly reversible, degeneration of articular cartilage induce tissue softening, generally preceding fibrillation and, thus, visible changes of the cartilage surface. We have already developed an indentation instrument for measuring arthroscopic stiffness of cartilage with typical thickness >2 mm. The aim of this study was to extend the applicability of the instrument for the measurement of thin (<2 mm) cartilage stiffness. Variations in cartilage thickness, which will not be known during arthroscopy, can nonetheless affect the indentation measurement, and therefore optimization of the indenter dimensions is necessary. First, we used theoretical and finite element models to compare plane-ended and spherical-ended indenters and, then, altered the dimensions to determine the optimal indenter for thin cartilage measurements. Finally, we experimentally validated the optimized indenter using bovine humeral head cartilage. Reference unconfined compression measurements were carried out with a material testing device. The spherical-ended indenter was more insensitive to the alterations in cartilage thickness (20% versus 39% in the thickness range 1.5-5 mm) than the plane-ended indenter. For thin cartilage, the optimal dimensions for the spherical-ended indenter were 0.5 mm for diameter and 0.1 mm for height. The experimental stiffness measurements with this indenter correlated well with the reference measurements (r = 0.811, n = 31, p < 0.0001) in the cartilage thickness range 0.7-1.8 mm. We conclude that the optimized indenter is reliable and well suited for the measurement of thin cartilage stiffness.