This study reports on early results following the introduction of one measure of medication prescription errors, that being the prescribing of a drug for which there is an 'alert' notice, into The Australian Council on Healthcare Standards Accreditation process. Characteristics of hospitals reporting of zero and non-zero errors were analysed using a logistic model. After adjusting for other hospital characteristics and duration of data collection, hospitals over 100 beds were more likely to report medication errors compared to hospitals with 1-100 beds. Reporting of these prescribing errors was not associated with the particular type or location of the hospital. However, as a result of monitoring of this indicator, a number of hospitals reported an increase in their quality assurance activities. It is a sentinel event and not a rate based indicator and, as a performance measure, is of greater value as an internal, rather than external, review mechanism.