Most studies investigating the relationship between plant diversity and ecosystem functioning lasted only a few years. These studies generally showed a positive relationship between diversity and productivity that strengthened with time. This pattern suggests the experimental communities have not yet reached maturity, which raises the question whether a positive relationship between diversity and productivity will remain if communities approach equilibrium conditions. Here, we analyze eight years of data from a plant diversity experiment without legumes. A positive relationship between plant diversity and productivity arose in the second year and increased until the fifth year. This relationship persisted throughout the remainder of the experiment, but the slope of the relationship remained constant at approximately 50 g/m2/log2[richness]. The positive relationship between diversity and productivity was caused by strong complementarity effects, which initially increased but remained constant after the fourth year of the experiment. Our results show a positive relationship between diversity and productivity is a long-term phenomenon, driven by complementary interactions.