Rice pest control is a critical challenge in the agricultural sector that requires a deep understanding of rice pest management. Regression analysis is a statistical method capable of describing and predicting cause-and-effect relationships between individuals. In real-life applications, not all relationships exhibit a known curve pattern, and non-identifiable curve forms are often observed. Additionally, a single cause may affect more than one outcome, and the outcomes themselves can have interrelationships. Such relationships can be approached through a multi-response semiparametric regression using a truncated spline multi-group model. This study aims to develop a multi-response semiparametric multi-group regression model using the truncated spline approach to understand the variables influencing rice pest control under light and dark conditions. This model is applied to secondary and simulated data with various scenarios to determine the best model. The study results indicate that the optimal model for secondary data is a semiparametric regression model with a linear order and a single knot point, achieving a determination coefficient of 89.17%. Simulation results show that the scenario 1 model (linear with a single knot point) produces a high determination coefficient. This multi-response regression model proves more optimal when error variance and multicollinearity levels are kept low to moderate.