This paper presents the application of an innovative wind-scaling method based on a power law to facilitate the modeling of extreme waves in mountain lakes. This power law is derived using long-term meteorological observations from numerous measuring stations. A reference WINDSIM wind field is calculated for each wind situation and extrapolated to extreme wind events (return periods up to 300 years) using a power law and statistics from a representative meteorological station. The power law scaled wind data and the wave data computed with the SWAN model are compared with results obtained using linear scaling for the case study of Lake Lucerne. This lake, located in the center of Switzerland, has a very characteristic shape and three dominant winds (Foetal, Bise and West Wind), which provide extremely heterogeneous wind and wave patterns. Results obtained using the new methodology yield improved wave predictions and can inform the design of shoreline prevention measures against extreme events.