An improved method to model CIP plants by plant data is presented. The Turkish gold cyanide leaching CIP plants, Bergama Ovacik, Gumushane Mastra and Eskisehir Kaymaz were modeled by using gold concentrations of monthly carbon, solution and solids spot samples of adsorption tanks without a need of laboratory work. Five carbon adsorption models were used namely k-n, Nicol-Fleming, Dixon, Film Diffusion with Langmuir isotherm and Film Diffusion with Freundlich isotherm (Johns model). Several monthly obtained plant data were collected and modeled separately and model parameters as well as regression coefficient R² values were found by non-linear regression. By comparing R² values, the best fit model for the three CIP plants was determined as Film Diffusion with Langmuir Isotherm that R2 values were above 0.95. Thus, by using the best fit model, the Kaymaz plant was optimized for existing 49 Mg per hour ore feed tonnage and plant parameters were predicted according to planned feed tonnage increase to 120 Mg per hour.