The stiffness and thickness distribution of healthy lumbar apophyseal cartilage was measured in 25 lumbar motion segments (L1–4). The cartilage indentation and needling techniques of Swann and Seedhom (37) were suitably altered to cope with the low modulus and small size of the joint surfaces. A load of 3.12 ± 0.19 N (mean ± s.d.) was applied to the cartilage through a hemispherical indenter of 4.756 mm diameter. The stiffness was calculated using the displacement and instantaneous load 150 ms and 2 s after the indenter first contacted the surface, and using the equations of both Waters (46) and Hayes et al. (47). The mean stiffness of apophyseal joint cartilage was 2.8 M Pa ± 4 per cent (mean ± 95 per cent confidence limit), and thickness 1.02 mm ± 3 per cent. Peripheral apophyseal joint cartilage was softer than central cartilage. The stiffnesses of the centres of the superior and inferior joint surfaces were 3.01 MPa ± 12 per cent and 3.55 MPa ± 11 per cent. Inferior surfaces had a mean thickness of 0.93 mm ± 5 per cent and stiffness of 2.88 MPa ± 7 per cent. Superior surfaces had a mean thickness of 1.10 mm ± 4 per cent and stiffness of 2.74 MPa ± 5 per cent. It was found that the stiffness of cartilage calculated according to the formula of Hayes et al. (47), based on a constitutive analysis of a thin isotropic elastic layer, was directly proportional to both the stiffness calculated using the semi-empirical formula of Waters (46), derived to describe indentation of thin rubber sheets, and the nominal compressive creep modulus calculated by dividing the mean contact stress by the strain at the deepest point of indentation. The creep modulus calculated 2 s after contact was directly proportional to the creep modulus calculated 150 ms after indenter contact, implying that deformation behaviour was uniform between these points despite variation of cartilage stiffness.