OBJECTIVES:Spanish Health Administration has developed some regulatory actions to check the growth of drugs expenditure. Specifically, in 1999, Government approved a reference price system for many drug groups (homogeneous). Here, we present the initial data from a study of drug consumption and a estimation of the saving with the prescriptions of generic drugs.METHODS: This study was divided into two periods: 1990- 1998 and 1999-2002. In the second, the influence of the reference price system implementation will be observed. Data from Spanish NHS about drug consumption were provided by the Ministry of Health and Consumer Affairs. We elaborated our own database. Consumption was expressed as PVP (price for sale direct to costumer, tax-free) by means of PTAM (Peseta Millions) or € M (Euro millions), and was revised through CPI (Consumption Price Indexes, Base 1990). Information about PVP drugs was obtained from Official Drug Directory. DHD (Dose for thousand inhabitants and for day) was also calculated according to DDD’s standard -Nordic Council Medicines ATC/DDD, edition 1998. Saving estimation was calculated through PVP minimum criteria of prescription, since the reference price system had not been implemented yet. RESULTS: According to consumption data obtained, in 1998, top four therapeutic groups - 29 active ingredients- explained almost 34 % of the whole expense - 286.116,2 PTA M or 2.458,33 € M. Besides, if the criterion of PVP minimum was applied, for the most usual format (package) per prescription, the average saving would be almost 7 %. In the Spanish pharmaceutical market there are many prices for specific active ingredients, being possible to identify the biggest individual savings. CONCLUSIONS: Almost the third part of NHS pharmaceutical budget in 1998 was assigned at 29 active ingredients (four therapeutical groups). The level of individual saving was significant in Nifedipine (47 %), Omeprazole (44 %) or Famotidine (42 %). Savings depend, specially, on the variation of dosage, presentation and packaging, among others.