Non-linear constraints in optimization problems can complicate the solution process by making the feasible region non-convex, leading to multiple local optima instead of a single global optimum. This can increase computational complexity and require specialized algorithms to find solutions. Additionally, non-linear constraints may limit the applicability of standard linear programming techniques, necessitating the use of non-linear programming methods.