In this article, the writers present a linear programming model by Freed andGlover (Freed and Glover, 1986) for discriminant analysis calculated with MicrosoftExcel program. The strong point of linear programming for discriminant analysis is thatit can avoid assumptions of the parametric statistics approach, such as data foranalyzing from multivariate normal distributions, or equal covariance matrices ofindependent variables between groups, or normal distribution of residuals. There are5 parts in this article: 1) prelude, which contains the general characteristics ofdiscriminant analysis, 2) description of the concept of linear programming for discriminantanalysis, 3) calculation of the linear programming using Microsoft Excel program, whichbegins with the first step and proceeds until arriving at the optimal results in themodel, 4) interpretation of calculations, and 5) conclusion in which the writers givesome suggestions for using this method.