The concept of pollinator niche complementarity maintains that species‐rich pollinator communities can provide higher and more stable pollination services than species‐poor communities, due to contrasting spatial and/or temporal pollination activity among groups of pollinators. Complementarity has usually been examined in pollinators’ patterns of flower visitation or abundance, while largely neglecting the possibility of complementarity in patterns of single‐visit contribution to fruit/seed set (pollination efficiency). However, variability in pollination efficiency can greatly affect pollinators’ overall pollination services and may therefore contribute an additional, important aspect of complementarity. In this study, we investigated the existence of pollinator complementarity in both visitation rates and pollination efficiencies. The study was conducted in 43 watermelon fields cultivated for seed consumption in a Mediterranean agro‐natural landscape in central Israel. We studied spatiotemporal variation in pollinators’ visitation activity, measured by repeated observations and netting, and single‐visit pollination efficiency, measured by the fruit and seed set rates of hermaphrodite flowers exposed to a single bee visit. Visitation and pollination efficiency were measured throughout the day and season, within and between fields with contrasting availability of nearby wild plants, and among flowers of different sizes. Pollinator species’ visitation rates as well as single‐visit fruit set efficiencies, but not seed set efficiencies, exhibited significant spatiotemporal variation that contributed to their complementarity. Pollinators’ visit frequencies were affected by surrounding land use, location within field, time throughout the season, and time of day. Pollinators’ fruit set efficiencies were affected by ovary size and time of day. Synthesis and applications. Crop pollinators may exhibit complementarity in both their visitation rates and pollination efficiencies, which can promote the overall level and stability of their pollination services. Complementarity in pollination efficiencies suggests further diversity effects on crop yield, and calls for taking into account the variability in pollination efficiency along spatiotemporal scales rather than considering it a constant, species‐specific trait. However, some modes of niche complementarity may not necessarily translate into increased pollination services and crop yield; the relevance and limitations of such mechanisms should be considered in the light of the specific crop and management system studied.