RESUMO
With the aims to evaluate the use of the linear regression analysis as a technique to detect interespecifc morphometric differences and to develop algorithms for differentiating oocysts from the Eimeria species, fecal samples of Santa Inês sheep, has been used. Empiric algorithms have been developed based on quantitative and qualitative data from the oocysts morphology. The linear regression of the length and width of the oocysts has been tested. There was no significant difference among the measures from the oocysts of the species E. bakuensis, E. faurei and E. ovinoidalis in lambs and in adults. From the five polar cap provided species, E. ahsata was differentiated from E. intricata and from E. bakuensis based on the straight lines angular coefficient. Concerning the species that do not present polar cap, E. parva and E. pallida were more subject to confusion. Even so, with the application of the algorithm, it was possible to cluster them into different strips. The efficiency presented by the algorithm for gathering these species was 77% and 64%, respectively. Regarding to the species with polar cap, the lowest efficiency presented by the algorithm was to cluster E. bakuensis, though it was above 50%.