BACKGROUND This study evaluated the ability of laser scanning tomography to distinguish between normal and glaucomatous optic nerve heads, and between glaucomatous subjects with and without field loss. METHODS 57 subjects were classified into three diagnostic groups: subjects with elevated intraocular pressure, normal optic nerve heads, and normal visual fields (n = 10); subjects with glaucomatous optic neuropathy and normal visual fields (n = 30); and subjects with glaucomatous optic neuropathy and repeatable visual field abnormality (n = 17). Three 10 degrees image series were acquired on each subject using the Heidelberg retina tomograph (HRT). From the 14 HRT stereometric variables, three were selected a priori for evaluation: (1) volume above reference (neuroretinal rim volume), (2) third moment in contour (cup shape), and (3) height variation contour (variation in relative nerve fibre layer height at the disc margin). Data were analysed using analysis of covariance, with age as the covariate. RESULTS Volume above reference, third moment in contour, and mean height contour were significantly different between each of the three diagnostic groups (p < 0.001). Height variation contour showed no significant difference among the three diagnostic groups (p = 0.906). CONCLUSIONS The HRT variables measuring rim volume, cup shape, and mean nerve fibre layer height distinguished between (1) subjects with elevated intraocular pressures and normal nerve heads, and glaucomatous optic nerve heads, and (2) glaucomatous optic nerve heads with and without repeatable visual field abnormality. This study did not directly assess the ability of the HRT to identify patients at risk of developing glaucoma. It is hypothesised that the greatest potential benefit of laser scanning tomography will be in the documentation of change within an individual over time.