In November 1990 major patient co-payment changes were introduced into the Pharmaceutical Benefits Scheme (PBS), which accounts for around 90% of all community prescriptions in Australia. Interrupted time series analyses were performed to assess the impact of these co-payment changes on the prescription levels of defined groups of 'discretionary' and 'essential' drugs for both the general community and for a subgroup comprising elderly returned servicemen and women. While the co-payment changes themselves had a significant immediate effect on lowering the use of both categories of drugs, the effects were substantially larger for the 'discretionary' groups in both cases. Notably there was a clear post-intervention trend for increased prescriptions of 'essential' drugs after the initial decline, which was not evident for the 'discretionary' drugs. The introduction of programmes to compensate high risk groups in Australia may have enabled the co-payment to become a more selective policy instrument than has been shown in other settings.