Multiplicative programming is a fundamental mathematical optimization problem in which the objective function contains a product of several real-valued functions. This paper deals with a class of multiplicative programming, called generalized linear multiplicative programming (GLMP), in which the objective is to minimize the product of two positive linear functions with general positive powers under linear constraints. Since the objective is a typical non-convex function, GLMP may have multiple local minima, making it computationally challenging. To address this, we propose a multi-objective optimization-based approach. By treating each function as an objective to be minimized, we show that a solution of GLMP is necessarily a non-dominated extreme point located on the vertex of the convex hull of the Pareto front. Then, we use a recursive algorithm to determine the set of all non-dominated extreme points. Notice that the solutions of GLMP can be directly extracted from this set. Furthermore, based on the Weighted Sum Method, it requires only solving one linear program in each iteration. Finally, we provide computational results on a specific instance of GLMP with 0–1 knapsack constraints, indicating that our approach is promising.