The species–area relationship (SAR) provides the foundation for much of theoretical ecology and conservation practice. However, by ignoring time the SAR offers an incomplete model for biodiversity dynamics. We used long-term data from permanent plots in Kansas grasslands, USA, to show that the increase in the number of species found with increasing periods of observation takes the same power-law form as the SAR. A statistical model including time, area, and their interaction explains 98% of variation in mean species number and demonstrates that while the effect of time depends on area, and vice versa, time has strong effects on species number even at relatively broad spatial scales. Our results suggest equivalence of underlying processes in space and time and raise questions about the diversity estimates currently used by basic researchers and conservation practitioners.