Abstract: Casing string combination design in deep wells and ultra-deep oil and gas wells, has inherent problems, which are nonlinear, multi-objective and combinatorial with discrete variables. This paper presents a method for design optimization of casing string based on a heuristic algorithm. The casing string combination was represented by a tree structure and parameterized, the design requirement vector reflecting the cost, weight and safety factors of casing string was given, and the adjustable multi-objective mathematical model based on this vector was established. A heuristic algorithm was proposed to solve the model. The analysis involved the comparison and analysis with the typical cases of casing string design. Considering that, the mathematical model established in this paper was used to obtain a solution that had better value than the original method in terms of fitness in the case of the most economic program. For example, solution 3 led to a reduction of cost and weight, and solution 2 led to a reduction in cost by 4.057%, but in increase in weight by 0.274%. The results show that the solution obtained by this method is reasonable and better than the original solution to some extent. This approach can effectively solve the complex casing string combination design problem in deep and ultra-deep wells.