The paper introduces Clifford Group Equivariant Simplicial Message Passing Networks (CSMPNs), a novel method for E(n)-equivariant message passing on simplicial complexes. CSMPNs integrate the expressiveness of Clifford group-equivariant layers with the topological complexity of simplicial message passing, which is more intricate than regular graph message passing. The method leverages Clifford algebras to represent geometric features such as areas and volumes through geometric products of vertices. To achieve efficient message passing, parameters are shared across different simplex dimensions, and the final message is an aggregation of incoming messages, termed *shared* simplicial message passing. Experimental results show that CSMPNs outperform both equivariant and simplicial graph neural networks on various geometric tasks, including convex hull volume prediction, human motion prediction, molecular motion prediction, and NBA player trajectory prediction. The implementation is available on GitHub.The paper introduces Clifford Group Equivariant Simplicial Message Passing Networks (CSMPNs), a novel method for E(n)-equivariant message passing on simplicial complexes. CSMPNs integrate the expressiveness of Clifford group-equivariant layers with the topological complexity of simplicial message passing, which is more intricate than regular graph message passing. The method leverages Clifford algebras to represent geometric features such as areas and volumes through geometric products of vertices. To achieve efficient message passing, parameters are shared across different simplex dimensions, and the final message is an aggregation of incoming messages, termed *shared* simplicial message passing. Experimental results show that CSMPNs outperform both equivariant and simplicial graph neural networks on various geometric tasks, including convex hull volume prediction, human motion prediction, molecular motion prediction, and NBA player trajectory prediction. The implementation is available on GitHub.