<p>This paper examines the issue of how the central government should allocate local government debt limits in China. The maximum default loss from local government debt that the central government can cope with is used as a proxy for its crisis disposal capability. Given the central government's limited ability to dispose of local debt crises, this paper draws on credit risk research to establish a quantitative relationship between the default probability of each local government's debt. It then uses the KMV model to determine the relationship between a local government's debt limit and its default probability. The default probability is treated as the decision variable in the optimization problem. The objective function aims to minimize the difference between the "sum of allocated local government limits" and the "total debt limit set by the central government." An optimization model is then established for the allocation of local government debt limits. Using 2021 as an example, the results from solving the optimization model with the genetic algorithm show that the calculated debt limits for six local governments, including Beijing and Shanxi, exceed the actual allocated limits, while the calculated limits for other local governments fall below the allocated limits (i.e., under-allocated). Notably, the over-allocations in Henan, Hebei, and Guizhou are particularly significant. Compared to the current debt limit allocation method, this approach is more effective in assisting the central government in preventing and controlling local government risks.</p>