This paper presents a systematic approach to generate walking primitive databases for anthropomorphic 3D-bipeds allowing step length adaptation, direction changes and stepping over obstacles. The individual walking primitives are derived by optimal control techniques. Zero moment point (ZMP) and friction conditions at the feet ensuring postural stability of the biped, as well as bounds on the joint angles and on the control torques, are treated as constraints. The resulting reference trajectories are validated by dynamic simulations.