The aim of this study was investigation optimum conditions of four Iranian commercial pistachio kernels (Fandoghi, Ahmad Aghaei, Akbari and Kale Ghoochi cultivars) infrared ROASTING. The ROASTING process was optimized by response surface methodology. Optimization was done by central composite design for every cultivar separately. The independent variables were infrared lamp voltage (V) and sample-lamp distance (m). The responses were chosen based on different aspects of quality factors: texture, color, taste, and time of ROASTING. For texture, the first and second fracture point; for color, browning index (BI); for taste, ROASTING taste; and for time, the ROASTING time to 2% moisture samples. The samples first fracture point were in range 20-40/5 N, second fracture point were in range 37-55 N, browning index were in range 38-41, ROASTING taste score were in range 4-9/12, and ROASTING time were in range 127-746 S, in the optimum points. A full quadratic model was fitted for responses by considering trial and error. The fitted model was significant based on ANOVA and R2. At last, results shown optimized points were different for every cultivar. The fitted model and experimental responses had a good correlation ( P ≤ 0. 05). Considering final products quality, it seems that pistachio kernels IR ROASTING can be a useful method to pistachio nut kernel ROASTING and can reduce the process time, at least 50%.