Abstract
Parasuicide rates have shown a marked increase in recent years and now consume a substantial proportion of medical services. To reduce hospital admission rates for parasuicide is probably desirable. While there has been considerable research on parasuicide, it is apparent that many of the findings have limited application to prevention. It is argued that attempts at primary prevention should be directed at those factors in the population considered to have a substantial function in promoting this group of behaviours. The first step in such a program should be the empirical delineation of a typology of parasuicide, since it is not a unitary phenomenon and a single preventive measure is therefore unlikely to have widespread success. Following the recognition of different requirements for different population groups, an experiment in prevention might be attempted by monitoring the parasuicide rates in two closely similar populations, one of which has alterations made sequentially to variables such as the following: the availability of tablets and the form of their presentation; the availability of personal advisory services; and community attitudes to parasuicide. Possible dangers inherent in such experimental epidemiology are acknowledged but it is argued thai aspects of the current approach to the problem of parasuicide may be in need of revision.