Introduction. Coffee's peel is the main sub product of coffee industry in Colombia and is not sufficiently used to exploit its bioactive potential, represented by its carotenoids, gallic acid, anthocyanins and antioxidants, all of them with great value for food and pharmaceutical industries. It is, therefore, important to study its granulometry in order to develop extraction processes. Objective. Analyze and make a model of coffee's peel granulometry (Coffea arabica L.), Castillo variety. Materials and methods. The raw material used in this research work was coffee peel (Coffea arabica L.), Castillo variety, acquired in farms from the Antioquian South West region. The coffee plants had an average of ten years of age and the selection was made in the 2015 harvest. The drying and the adequacy of the raw material took place in one of the farms, in Caldas, Antioquia, Colombia. The granulometry analysis by sieving was performed at Corporación Universitaria Lasallista's laboratories. A cutter mill was used and a grinding time lapse between 1 and 4 minutes was assessed. The particles distribution was studied with the Gates-Gaudin-Schuhmann (G-G-S) and Rosin-Rammler (R-R) models. Both models were linearized in order to obtain the equation constants and the correlation coefficients to be compared. Results. The Rosin-Rammler's distribution function is the one that best represents the granulometric distribution of coffee peel, with a r2 equal to 0,99 in contrast to the distribution function of Gates-Gaudin-Schuhmann, with a lower r2 that decreases as grinding time increases. [ABSTRACT FROM AUTHOR]