The extraction of phenolic compounds from Inga edulis leaves was optimized by the simultaneous maximization of the yield in total phenolics (TP), total flavanoids (TFA), and total flavonols (TFO), using the response surface methodology (RSM). A first set of experiments allowed identifying the temperature, the time of contact and the ethanol proportion in the extraction solution, as the main variables affecting the extraction efficiency. A rotatable central composite design consisting of 18 experimental runs with three replicates at the centre point was then applied and a second-order polynomial model was used to describe the experimental data regarding TP, TFA and TFO. The experimental results fifted well to the model and more than 85% of the variability was explained. TP, TFA and TFO showed different patterns of extractability, with significant variation in the linear. quadratic, and interaction effects of the independent variables. The optimized conditions were 86.8% ethanol, 58.2 degrees C and a time of contact of 46.8 min. The corresponding predicted values were 134.6 mg gallic acid equiv./g dry matter (DM), 26.0 mg catechin equiv./g DM and 13.8 ma rutin equiv./g DM, for TP, TFA and TFO, respectively. The experimental values agreed with those predicted within a 95% confidence interval, thus indicating the suitability of RSM in optimizing the extraction of phenolics from L edulis. (c) 2007 Elsevier B.V. All rights reserved.