Statistics based experimental design on chitinase production by Trichoderma harzianum was optimized in solid state fermentation using Plackett-Burman design and response surface methodology. The important medium components identified by initial screening method of Plackett-Burman were peptone, malt extract, citric acid and urea. Plackett-Burman Pareto chart illustrates the order of significance of the variables affecting the cellmass production. Central composite response surface methodology was performed to evaluate the effects of temperature, pH, inoculum size and substrate concentration on production of chitinase by Trichoderma harzianum using sugarcane bagasse under solid state fermentation. Statistical analysis of results showed that, the linear and quadric terms of these four variables had significant effects and evident interactions existing between pH and inoculum size were found to contribute to the response at a significant level. After optimization, the maximum enzyme yield was 34 U/mL from agroindustrial waste sugarcane bagasse.