Coral reef restoration aims to help threatened coral ecosystems recover from recent severe declines. Here we address whether coral fragments should be out-planted individually or in larger aggregations. Theory suggests alternative possible outcomes: whereas out-plants within aggregations might suffer from heightened negative interactions with neighbors (e.g. competition for space), they may alternatively benefit from positive interactions with neighbors (e.g. buffering wave disturbances). On a degraded reef in the Caribbean (St. Croix, USVI), using out-plants of the critically endangered staghorn coral Acropora cervicornis, we experimentally tested how aggregation density (1–20 out-planted coral fragments spaced at approximately 5 cm) influenced initial coral growth (over 3 months). Coral growth declined as a function of aggregation size, and out-plants within larger aggregations had fewer and shorter secondary branches on average, indicative of horizontal competition for space. Our results therefore suggest that wide spacing of individuals will maximize the initial growth of out-planted branching corals.