In this work, efficient methods to determine aluminum (Al) and lead (Pb) in beer samples by graphite furnace atomic absorption spectrometry (GFAAS), following multivariate optimization, were developed. The only sample preparation used was degassing of the beer to remove CO 2. Permanent rhodium (Rh) for aluminum and permanent iridium (Ir) for lead proved to be the better permanent modifier. Optimization included fractional factorial planning using the Pareto and the CCD designs. The working linear range was 0-90 µg L-1 for aluminum and 0-60 µg L-1 for lead (r 2 >0.99). The obtained LOQs were 4.8 and 1.6 µg L-1 for aluminum and lead, respectively. Aqueous and matrix-matched calibration curves had average angular coefficients that were not statistically different, i.e., the matrix effect was absent for aluminum, while for lead the coefficients were statistically different. Matrix-matched calibration was used in the subsequent studies for the beer samples. The accuracy was checked by recovery tests with a percentage recovery range of 89-95% (n=45) for Al and 97 to 98% for Pb (n=45). The coefficient of variation (CV) of the intra-assay studies (n=15) was 4.0 and 2.5% for Al and Pb, respectively, where the CV for the inter-assay studies (n=45) was 3.0 for Al and 3.1% for Pb. The Al and Pb levels in the 21 beer samples varied from