It has been suggested that over-reliance on the cup-to-disc ratio is a major factor in the misinterpretation of the optic disc. In spite of this optometrist assessment of the optic disc tends to be restricted to measurement of the cup-to-disc ratio and cup depth only. Would interpretation of the disc improve if optometrists were to evaluate other parameters? The aim of this study was to evaluate the accuracy of optometrist assessment of nine parameters of the optic nerve head using direct ophthalmoscopy.Eight optometrists evaluated nine parameters of the optic nerve head (vertical disc diameter, vertical cup-to-disc ratio, neuroretinal rim configuration, cup shape, neuroretinal rim colour, vessel path, presence/ absence of haemorrhage, extent and location of peripapillary atrophy and classification of health status of the disc) in 50 eyes of 50 patients using direct ophthalmoscopy. Intensive training in optic nerve head assessment was given prior to assessing the patients. Criteria for evaluation were discussed. The 'gold standard' reference was the classification of the parameters by a consultant ophthalmologist with a special interest in glaucoma.Interobserver agreement for vertical cup-to-disc ratio was almost perfect (mean weighted kappa 0.84). Agreement for neuroretinal rim configuration, cup shape, haemorrhage and final classification of the disc was good (mean kappa 0.62-0.67). There was moderate agreement for vessel configuration (mean kappa 0.53). For assessment of peripapillary atrophy, disc size and neuroretinal rim colour, agreement was fair (mean kappa 0.22-0.34).Accuracy of assessment was greatest for vertical cup-to-disc ratio, neuroretinal rim configuration and cup shape. Improved agreement has been demonstrated for the final classification of the disc compared with previous reports. The combination of training and assessment of additional disc parameters appears to improve interpretation of the optic nerve head by optometrists.