Biodegradable implants have the potential to serve as next-generation temporary medical devices as they can safely dissolve in the human body upon bone regeneration. Degradation uncertainty of the biodegradable material can remarkably affect the mechanical performance of biodegradable composite implant structures. It is necessary to consider this issue when designing resorbable metallic composite structures. To this end, this study introduces a novel robust topology optimization approach for designing biodegradable composite structures considering the degradation rate uncertainty of the biomaterial. The density-based topology optimization method is used to track the evolving of biomaterial layout during the optimization process, the Expansion Optimal Linear Estimation (EOLE) method is used to model the degradation uncertainties, and the Polynomial Chaos Expansion (PCE) based uncertain propagation analysis is implemented to predict the stochastic response. Then the robust topology optimization problem is formulated, in which a weighted function that achieves a trade-off between the expected mean and standard deviation of the performance function of interest was used as the objective function. The sensitivities of the design variables were deduced by considering the material degradation over time. Several numerical examples were presented to demonstrate that the proposed method could generate meaningful optimal topologies with the desired mechanical performance.