The eggbeater kick is an important skill in artistic swimming necessary to lift the body above water level. Previous attempts to model its performance included complex biomechanical parameters that cannot be easily used to guide strength and conditioning training. The objective of this study was to model the relationship between hip strength and eggbeater performance through a machine learning algorithm. We assessed hip function of 92 elite artistic swimmers with six easily performed isometric tests. These data were fed to a gradient boosting model to predict three technical variables: body boost height [BB-H], eggbeater height [EB-H] and eggbeater force [EB-F]. Group mean differences () between predicted and measured variables were reported. Then, the model was used to propose training tips for two hypothetical case studies. Our model predicted performances with errors within the resolution of the scale used during competitions: absolute error of and in EB-H and BB-H, respectively. The predicted performance was similar to the measured one for all technical tests (EB-F: ; EB-H: ; BB-H: ). We illustrated some of the important predictors (hip internal rotation, abduction, and left-right imbalances) of the eggbeater kick performance and highlighted personalized strategies to improve performance.