Pilling of fabrics is a phenomenon caused by conglomeration of fibers and their accumulation on the surface of fabrics. This free mass of fibers seen on the surface of fabrics is called "pill". Increase in the number of pills on fabrics not only creates unpleasant appearance but also results in the abrasion of clothes and its early erosion. With washing or wearing of the clothes with each other pills are formed through the engagement of the fibers protruding from the fabric surface. Therefore, identifying the factors effective on pilling, as well as their measurement and control is very important. Among these factors length, linear density, tear strength, abrasion resistance, fibers count, yarn twist and structure of the fabric can be mentioned. The purpose of this study is to examine the effects of various fabric parameters such as material, yarn count and twist, structure and density of fabric as well as examination of the effect of feeding two parallel yarns into a flat knitting machine on the rate of pilling in weft knitted fabrics with blended wool/acrylic fibers. Therefore, three types of yarns with blended ratio of 25/75, 50/50 and 75/25 with grades of 10 and 15 metric yarn count and twist metric factor of 80 and 90 were used as samples. These yarns were knitted by a gauge 12 flat knitting machine in three structures: single weft knitted, simple rib, and full Milano with both low and high densities. After preparing the samples, they were exposed to erosion with Martindale machine with125, 500, 1000, 1500 and 2000 rotations. Then, the formed pills were counted in three ways: eye counting, photography in a dark room and scanning of surface of sample. Afterwards, the resulted data were analyzed by using Taguchi experiment design program based on which the samples had been produced; and the optimum conditions of different fabric factors were obtained. Results of this study showed that among the parameters which used in design of the experiments, yarn material, density and structure type of fabric (in sequence) had the highest effect on formation rates of pills. Therefore, the lowest numbers of pills after 2000 cycles of abrasion on yarns knitted with 75 percent of acrylic were formed with full Milano structure. Also according to these results, with increasing density and yarn twist, the number of pills formed on fabrics declines. Finally, these results were used to estimate the number of pills in these types of yarns through neural network method. Key words : wool/acrylic blend fibers, abrasion resistance, pill, weft knitted fabric, Taguchi method, neural network