In the last few years, bioactive components or their extraction techniques are gaining special interest in scientific areas. In this framework, orange leaves were used for preparation of extracts with high content of biologically active compounds. To optimize the extraction process, three levels and three variables of Box–Behnken design with response surface methodology were applied. Investigated responses were the total phenolic content (TPC), 1,1-diphenyl-2-picrylhydrazyl (DPPH) radical, cupric ion reducing antioxidant capacity (CUPRAC), and ferric reducing antioxidant power (FRAP). Independent variables were methanol concentration (10–90%), temperature (20–60°C), and extraction time (60–180 min). Experimentally obtained results were fit into a second-order polynomial model with multiple regression. Analysis of variance was used to estimate model fitness and determine optimal conditions for processing. Estimated optimal conditions were 90% methanolic solution, 60°C and 180 min using these parameters; the predicted values of investigated responses were 43.19 mg GAE/g (GAE: gallic acid equivalents), 43.04 mg TE/g (TE: trolox equivalents), 139.34 and 93.76 mg TE/g for TPC, DPPH, CUPRAC, and FRAP, respectively. The obtained optimal conditions could be considered as an alternative strategy for developing novel functional products.