Randomized full-factorial designs (3(2)) were used to investigate the effects of processing conditions in the top-spray fluidized bed (TSFB) on the granulation of acetaminophen powder (USP) using 5% polyvinylpyrrolidone (w/w) as the binder. Measured granule properties included the following: mean size and size distribution, specific surface area, bulk density, tapped density, flow rate through an orifice, angle of repose, residual moisture content, and percent overs (> 2 mm). The granules were then compressed (500, 1000, 1500 lbs) into tablets (9-mm shallow concave) using an instrumented rotary press and analyzed for both physical properties and drug-release characteristics. All experimental batches were run in triplicate to reduce the possibility of erroneous results and to increase the confidence in the resulting empirical relationships derived using response-surface methodology. Measured responses were then related to process parameters using two-factor and three-factor linear, interactions, and quadratic regression models. These models were used to generate three-dimensional response surfaces for use in the final analyses. Coefficients of determination (R2) ranging from 0.08 to 0.81 were obtained, indicating that only a portion of the variation in the data could be explained by the changes in process parameter settings during granulation and tableting. The best overall model fits were observed for mean granule size, size distribution, bulk density, tapped density, percent drug dissolution, tablet disintegration time, and tablet friability.