Surface response methodology was applied to the optimization of the laboratory scale production of tannase using a lineage of Paecilomyces variotii. A preliminary study was conducted to evaluate the effects of variables, including temperature ( degrees C), residue (%) (coffee husk:wheat bran), tannic acid (%) and salt solutions (%) on the production of tannase during 3, 5 and 7 days of fermentation. Among these variables, temperature, residues and tannic acid had significant effects on tannase production. The variables were optimized using surface response methodology. The best conditions for tannase production were: temperature (29-34 degrees C); tannic acid (8.5-14%); % residue (coffee husk:wheat bran 50:50) and incubation time of 5 days. The supplementation of external nitrogen and carbon sources at 0.4%, 0.8% and 1.2% concentration on tannase production were studied in the optimized medium. Three different nitrogen sources included yeast extract, ammonia nitrate and sodium nitrate along with carbon source (starch) were studied. Only ammonia nitrate showed a significant effect on tannase production. After the optimization process, the tannase activity increased 8.6-fold.