Response surface methodology (RSM) based on a three-factor and three-level Box–Behnken design was employed for optimizing the aqueous ultrasound-assisted extraction (AUAE) conditions, including extraction time (35–45 min), extraction temperature (45–55 °C) and ultrasonic power (150–250 W), for the recovery of total phenolic content (TPC) and rutin from lemon by-products. The independent variables and their values were selected on the basis of preliminary experiments, where the effects of five extraction parameters (particle size, extraction time and temperature, ultrasonic power and sample-to-solvent ratio) on TPC and rutin extraction yields were investigated. The yields of TPC and rutin were studied using a second-order polynomial equation. The optimum AUAE conditions for TPC were extraction time of 45 min, extraction temperature of 50 °C and ultrasonic power of 250 W with a predicted value of 18.10 ± 0.24 mg GAE/g dw, while the optimum AUAE conditions for rutin were extraction time of 35 min, extraction temperature of 48 °C and ultrasonic power of 150W with a predicted value of 3.20 ± 0.12 mg/g dw. The extracts obtained at the optimum AUAE conditions were compared with those obtained by a hot water and an organic solvent conventional extraction in terms of TPC, total flavonoid content (TF) and antioxidant capacity. The extracts obtained by AUAE had the same TPC, TF and ferric reducing antioxidant power as those achieved by organic solvent conventional extraction. However, hot water extraction led to extracts with the highest flavonoid content and antioxidant capacity. Scanning electron microscopy analysis showed that all the extraction methods led to cell damage to varying extents.