Fuzzy logic 1 has potential application in fields such as pattern recognition and process control. Since Liu first introduced an optical fuzzy logic processor utilizing a lens-array-based multiple imaging system, 2 many other systems have also been proposed and demonstrated. Most of early implementations were based on the principle of shadow-casting, with spatially encoded patterns being superimposed on each other by use of either light source array 3 or lens-array. 2 To obtain correct output of the fuzzy logic maximization (or minimization) operations, thresholding devices were needed in some systems. These thresholding devices, as well as the complex encoding patterns, make the systems complicated. Other systems utilized a complex encoding scheme which resulted in an output pattern different from the input patterns. Thus, the encoding scheme proposed for two-input fuzzy logic operations was difficult to be extended to multiple-input operations. 3,4, In this paper, we propose and demonstrate a novel optical fuzzy logic processor based on four-wave mixing in photorefractive crystals. Specifically, the recording of light-induced gratings is utilized to achieve minimization operations, while the readout of degenerated gratings is utilized to achieve maximization operations. Our system has several advantages including simple data encoding scheme, full parallelism, high speed, high accuracy, and simple architecture (no thresholding devices).