Talc is extensively used in a wide vaFiety of cosmetic products , particularly in powder products . Talc keeps the skin feeling smooth and dry. Talc in the pharmaceutical industry is used as a glidant and lubricant. Glidants such as talc improve flow properties of powder by decreasing interparticulate friction, by decreasing van der Waals forces and electrostatic charges, by changing particle size distribution, and by decreasing the effect of humidity on surfaces of host particles by forming a mechanical barrier. The loosely bound lattice layers slide over each other and form roller structures which explains its lubr ication characteristics. Talc has less deleterious effect compared to magnesium stearate on tablet in vitro properties. Talcs from different size grades were evaluated for their use in direct compression tablet formulations. Lubricant efficiencies of talcs were measured using ejection force values. The Supra grade of the Cyprus Industrial Mineral company talc was found to be most efficient of the grades tested as lubricant and also gave tablets of more acceptable in vitro properties than t ablets lubricated with magnesium stearate only. Additionally talcs in substantial percentages were evaluated for their potential as a direct compression matrix material. With commonly used excipients such as microcrystalline cellulose and lactose, the formulations were self lubricating and the tablets ejected easily. Tablets with very low friability, high crushing strength, rapid dissolution rates, good weight uniformity, content uniformity and potency of drugs were obtained. At similar compression forces, tablet hardness with the 300 grade of Alabama, Altalc, and Beaverwhite talcs was significantly greater than corresponding 400 and 500 grades. An optimum direct compression tablet formulation of a conventional theophylline tablet was achieved using the technique of response surface methodology and successive quadratic programming (SQP) . The response surfaces were obtained from a second order uniform precision hexagonal design. The tablet formulation was optimized for mean 1.n vitro dissolution time using friability, hardness, ejection force and disintegration time as constraints within the experimental region by the SQP algorithm. The response surface model was validated by preparing and evaluating the predicted formulation. The characteristics of the tablet formulation were analyzed by principal component analysis. Sensitivity analysis of optimal solution was performed for each constraint, while all remaining constraints were held constant. The robustness of the response surface model was evaluated by simulation for error in the compression force values due to its inherent variation. Although pharmaceutical scientists have previously reported optimization studies, the approach used in this thesis has signifi cant advantages and has not apparently been previously used for the optimization of pharmaceutical formulation. A completely novel use for talc as a major matrix component in direct compression tablet formulation, has been proposed and examined . A robust and efficient response surface experimental design and mathematical optimization technique has been evaluated for application to the pharmaceutical sciences . The computer search method has the disadvantage of frequently giving a plural solution for suitable formulations. With the classical Lagrangian method , it can become difficult to solve the resultant set of simultaneous equations especially for nonlinear problems. The SQP method is more efficient and robust compared to previously published optimization techniques including SUMT (Successive Unconstrained Minimization Technique) . The SQP method can also be used to obtain solution to the problems solved by the previously mentioned optimization methods. ACKNOWLEDGEMENTS I would like to offer my sincere gratitude to my major professor Dr. Christopher T . Rhodes for his advice , guidance and kind support . I greatly appreciate the opportunities Dr. Christopher T . Rhodes has created for me. Our association has progressively become more enjoyable and fruitfull . Dr. Christopher T . Rhodes kind support greatly helped me in finishing the dissertation in time . A special thank you for Dr. E.R. Suryanarayan, for his guidance and ready availibility for many helpful discussions. His pleasant and friendly help is greatly appreciated. I also thank Dr . Charles Woodruff, Dr. Hosein Zia, Dr. David Woodford , and Dr. G. Verma for their assistance, encouragment and guidance. I appreciate the guidance and assistance of Dr. William D. Lawing, Dr. Edward Carney, Dr. Gregory Peck on statistical analysis. I appreciate the advice and encouragement from all my friends. Also I would like to thank Nina Kajiji and the staff at academic computer center for their assistance with SAS programs. Last but not least, I would like to offer my thanks to my parents and family members for their constant support, encouragement and love .