The growing interest in sonochemistry as a tool for environmental remediation leads to the need for process optimization. Sonochemistry is a complex process, which depends on physical parameters and also on the process conditions. Physical parameters are interrelated and therefore a systematic approach has to be taken to optimize the process. The effect of physical parameters on the destruction of seven estrogen hormones (17α-estradiol, 17β-estradiol, estriol, 17α-ethinylestradiol, 17α-dihydroequilin, estrone and equilin) is reported in this study. Artificial neural networks (ANN) was used as a tool to identify the correlations between these process parameters. ANN enabled the establishment of relationship between sonication parameters such as power density, power intensity, ultrasound amplitude, as well as the reactor design parameters. The major significance was attributed to the area-specific power density and the volume-specific power intensity. The results of this work provide a sound basis to design pilot and full-scale ultrasound treatment systems. Process optimization lead to a 5-fold decrease in energy consumption as compared to the commercially available reactors, thereby making the process attractive for field applications.