This paper describes the strategy to optimize GTC telescope’s optical performance in terms of reflectivity and scattering by means of a suitable combination of mirror coating, CO<sup>2</sup> and in-situ cleaning. According to our experience, a monthly CO<sup>2</sup> cleaning was established, except during sporadic dust episodes, when a shorter weekly period is much more appropriate. Trends of the main optical parameters were recorded and analyzed to identify possible causes for the variation of the mirrors performance. As the total reflectivity stems from the combination of three optical surfaces, we set thresholds for the individual components and used these to select the mirrors that have to be replaced and cleaned. We also compared historical data about total reflectivity with optical OSIRIS zeropoints evolution and established a nonlinear relation, that is applicable in the periods where direct measurements on the mirror surface are not feasible. In this line, we are working on an innovative method to estimate the reflectivity for a segmented mirror based on the zeropoints measurement for the individual segments obtained by un-stacking the primary mirror under a controlled pattern.