OBJECTIVES: The Brazilian Extended Consumer Price Index (IPCA) for Pharmaceutical Products (IPCA-Pharmaceuticals) measures changes in the prices of a fixed basket of medicines purchased by Brazilian households. Monitoring, modeling and prediction of the index are important because adjustments for inflation must be made in accordance with the principles of good practice in health economic analyses. In addition, the index has an important weight in the calculation of the General IPCA, the official index to guide inflation-related policies, and is used to regulate pharmaceutical products prices. The objectives of this study are: 1) to model the IPCA-Pharmaceuticals time series during the sample period 2006-2010; and 2) predict the rate of inflation for the year 2011. METHODS: Two classical methodologies of time series analysis were employed: Holt-Winters exponential smoothing and the Box-Jenkins approach. Both methods were compared across three statistics based on errors measures: mean squared error, mean absolute percent error, and Theil's U coefficient. The best fitted model was chosen based on minimizing the error statistics and was used to forecast the IPCA-pharmaceuticals in 2011. RESULTS: Monthly IPCA-Pharmaceuticals data was collected from July 2006 to December 2010 (n=54) from Brazilian Institute of Geography and Statistics. The IPCA-Pharmaceuticals percentage change time series was converted to index number on base July 2006=100. The Holt-Winters additive method was adjusted and compared with a SARIMA (0,1,1)x(1,0,0)12 model estimated through the Box-Jenkins approach. The between-methods comparison showed a large advantage for the Box-Jenkins, which minimized the three errors measures. CONCLUSIONS: The Box-Jenkins method presented better results as compared to the Holt-Winters method. The final forecasting predicted a 2% inflation rate for pharmaceutical products by the end of 2011, which is lower than the general inflation target rate established by the Brazilian government (2.5-6.5%) and below the observed mean rate in the last three years (4.4%).