Eight studies found a robust negative relationship between the experience of power and the experience of loneliness. Dispositional power and loneliness were negatively correlated (Study 1). Experimental inductions established causality: we manipulated high versus low power through autobiographical essays, assignment to positions, or control over resources, and found that each manipulation showed that high versus low power decreased loneliness (Studies 2a–2c). We also demonstrated both that low power can increase loneliness and that high power can decrease loneliness by comparing these conditions to a baseline condition (Studies 3–4, 6). Furthermore, we establish a key mechanism that explains this effect, demonstrating that the need to belong mediates the effect of power on loneliness (Studies 5–6). These findings help explain some effects of power on social cognition, offer insights into organizational well-being and motivation, and speak to the fundamental question of whether it is “lonely at the top” or lonelier at the bottom.