ate for traditional oils (soybean, corn, canola, etc.) but not for tigernut oil, because tigernut meal contains relatively high levels of starch (up to 30%) , which is readily gelatinized and/or denatured 7,8) . Therefore, research on the separation technologies of tigernut oil has focused on coldpressing 9) and subcritical fluid extraction 5) . Mechanical cold pressing is easier to operate with many advantages over subcritical extraction, especially in retaining the natural compositions and flavor of the oil. However, the oil yield from tigernut is relatively low in cold-pressed extraction, which is potentially due to the partial denaturation of tigernut starch 10,11) . Because of its high yield and low risk of starch denaturation, subcritical fluid extraction is attracting increasing attention and was used for this study. For this extraction method, different low-boiling solvents, such as n-butane, n-propane, dimethyl ether and tetrafluo-Abstract: In this study, tigernut oil was extracted from tigernut meal by subcritical n-butane extraction with the assistance of microwave pretreatment. Effects of microwave pulse duration, particle size of tigernut meal, and subcritical extraction variables (temperature, time, solid-liquid ratio, number of extraction cycles) on extraction efficiency were examined by single-factor experiments and Response Surface Methodology (RSM) modeling. The results indicate that microwaving (560 W, 6 min) significantly increased the subcritical extraction efficiency. The variation of extraction yield could be interpreted as a nonlinear function of extraction time, temperature and liquid-solid ratio. Changing the independent variables could affect the oil extraction efficiency. The subcritical extraction of tigernut oil with a liquid-solid ratio of 3.62 kg/(kg of tigernut meal) at a temperature of 52℃ for 32 min after three extraction cycles produced the most oil, and a maximum yield (24.736%) of tigernut oil was achieved. The ratio of unsaturated to saturated fatty acids (4.68 UFA/SFA), low acid value (3.30 mg KOH/g oil), low peroxide value (0.28 meq.kg -1 ), and preponderance of oleic acid indicate a high-quality oil. To describe the extraction kinetics, a modified Brunner's mathematical model was used. The model fit the experimental data well over the entire operating range, and the explanation coefficient exceeds 96%. Our results can be used to develop an optimized method for subcritical fluid extraction of tigernut oil and can move industry further toward implementing microwave-assisted subcritical extraction in oil processing.