In 1962, Perley and Guze introduced objective criteria for the diagnosis of hysteria (Perley and Guze, 1962). These criteria were essentially a quantification of clinical observations which had appeared in the literature previously, notably in the work of Purtell, Robins and Cohen (1951). The Perley and Guze criteria for the diagnosis of hysteria offered several important advantages. First, they were derived from observations of the natural history of hysteria. They were straightforward and objective, suggesting that they could be used reliably by different clinicians in different places. Second, the criteria were accompanied by a follow-up study which indicated that diagnoses of hysteria made by means of the Perley-Guze criteria would be stable over a six to eight year period in 90 per cent. of cases. Put another way, these criteria selected a population homogeneous in prognosis. The stability of such a population over time is of particular importance. The original Perley-Guze paper and a further study by Gatfield and Guze (1962) both described the prognosis of patients selected by looser criteria. When conversion (pseudoneurologic) symptoms alone were the criteria of diagnosis, patients developed a bewildering array of psychiatric, medical and neurological illnesses within a few short years. In contrast, the Perley-Guze criteria for the diagnosis of hysteria predict prognosis accurately and represent a significant advance in descriptive, clinical psychiatry.