Linear discriminant analysis, a multivariate statistical procedure, applied to serum calcium, phosphate, alkaline phosphatase, bicarbonate, chloride, creatinine and tubular reabsorption of phosphate, proved to be effective in distinguishing patients with Primary Hyperparathyroidism from other hypercalcaemic patients in eithy-four retrospective cases. The application of the model to thirty-four prospective cases enabled us to separate correctly, hyperparathyroid patients from non-parathyroid hypercalcaemic patients.