WoCoCo (Whole-Body Control with Sequential Contacts) is a unified framework designed to learn whole-body humanoid control with sequential contacts. Traditional methods, such as model-based motion planning, are time-consuming and rely on simplified dynamics models, while model-free reinforcement learning (RL) offers robustness but often requires tedious task-specific tuning and suffers from long-horizon exploration issues. WoCoCo addresses these challenges by naturally decomposing tasks into separate contact stages, facilitating simple and general policy learning pipelines through task-agnostic rewards and sim-to-real designs. The framework enables end-to-end RL-based controllers to perform four challenging humanoid tasks involving diverse contact sequences in the real world: versatile parkour jumping, box loco-manipulation, dynamic clap-and-tap dancing, and cliffside climbing. Additionally, WoCoCo is applied to a 22-DoF dinosaur robot for loco-manipulation tasks, demonstrating its versatility and universality. The paper includes detailed reward designs, sim-to-real transfer methods, and case studies to showcase the effectiveness of WoCoCo in various dynamic tasks.WoCoCo (Whole-Body Control with Sequential Contacts) is a unified framework designed to learn whole-body humanoid control with sequential contacts. Traditional methods, such as model-based motion planning, are time-consuming and rely on simplified dynamics models, while model-free reinforcement learning (RL) offers robustness but often requires tedious task-specific tuning and suffers from long-horizon exploration issues. WoCoCo addresses these challenges by naturally decomposing tasks into separate contact stages, facilitating simple and general policy learning pipelines through task-agnostic rewards and sim-to-real designs. The framework enables end-to-end RL-based controllers to perform four challenging humanoid tasks involving diverse contact sequences in the real world: versatile parkour jumping, box loco-manipulation, dynamic clap-and-tap dancing, and cliffside climbing. Additionally, WoCoCo is applied to a 22-DoF dinosaur robot for loco-manipulation tasks, demonstrating its versatility and universality. The paper includes detailed reward designs, sim-to-real transfer methods, and case studies to showcase the effectiveness of WoCoCo in various dynamic tasks.