Factors governing the entry of new drugs into clinical practice are changing, with increasing emphasis on economic issues. In future, organisations that subsidise the use of pharmaceuticals are likely to require sponsors to provide evidence of the cost-effectiveness of their products. The first national government to signal such an intention is the Commonwealth Government of Australia, which from January 1993 will require economic analyses in support of applications for listing of new pharmaceutical products on its schedule of pharmaceutical benefits. This move is underpinned by legislation that requires the country's Pharmaceutical Benefits Advisory Committee (PBAC) to consider costs and effectiveness when recommending listing of new drugs. The approach that has been recommended to the Committee is based on advice from a group of consultants, health economists and clinicians. The PBAC will use economic analyses as an aid to decision-making that will remain within a clinical framework; the viewpoint will be societal, and analyses will include costs that fall outside the pharmaceutical benefits scheme. The preferred approach is comparative cost-effectiveness analysis with a particular emphasis on the marginal costs of obtaining additional health benefits with new drugs, compared with existing therapies. The use of analyses that are restricted to potential cost savings with new drugs is discouraged, as is the inclusion of indirect costs and benefits. To facilitate the conduct of economic analyses, it is planned to hold meetings with specialist clinicians to obtain consensus on a range of intermediate clinical outcome indicators, and to publish lists of 'standard' Australian costs that will be updated regularly. The approach being followed in Australia has implications for both the government and the pharmaceutical industry. The responsibility for monitoring the effects of this new policy lie with the government. The success, or otherwise, of the policy should not be gauged simply by the effects on the price of new drugs which, historically, have been relatively low in Australia. A full evaluation will require that more effort be put into clinical outcomes research and the development of population databases, an area in which Australia lags behind other countries.