VOODOO 3D is a high-fidelity 3D perceptual single-shot head reenactment technique. Our method transfers the driver's facial expressions to the source and produces view-consistent rendering for holographic displays. It is based on a fully volumetric neural decoupling framework for 3D perceptual single-shot head reenactment of source appearance and driver expression. Our method is real-time and produces high-fidelity, view-consistent output, suitable for 3D remote conferencing systems based on holographic displays. We demonstrate state-of-the-art performance on various datasets and showcase high-quality 3D perceptual head reenactment for challenging and diverse subjects, including non-frontal head poses and complex expressions of both source and driver.