The response surface methodology (RSM) was employed to optimize the microencapsulation condition of sunflower oil (SO) as a typical seed oil. The microencapsulation efficiency (MEE) of microencapsulated sunflower oil (MESO) was investigated with respect to four variables including SO concentration (X,), proportion of milk protein isolates (MPI) to coating wall (X-2), soy lecithin concentration (X-3), and homogenizing pressure (X-4). As a result, a polynomial regression model equation was fitted as follows: MEE (%) = 4.137772 + 3.524183X(1) + 3.475205X(2) + 2.914167X(3) - 0.074532X(2)(1) - 0.067482X(2)(2). Effect of homogenizing pressure was negligible. The optimal conditions for microencapsulation of SO were 23.6% SO, 19.0% MPI, 2.5% soy lecithin, and 54.8% dextrin, respectively. MESO under the optimized conditions gave rise to the highest MEE, approaching 96.6% of MEE. Compared to MESO showing a low MEE (70.2%), the peroxide value (POV) of the total oil from the MESO under the optimized conditions was significantly lowed even under the accelerated storage conditions at 60 +/- 1 degrees C after 30 days, which indicates a promising feature of RSM-mediated microencapsulation process of seed oil. (C) 2007 Elsevier Ltd. All rights reserved.