Regime shifts are usually attributed to when the magnitude of exogenous forcing gradually exceeds a threshold-value in a bistable system, resulting in a transition to a new self-reinforcing state. We argue that tipping points and thresholds in many ecosystems also need to be defined by the duration of exogenous pressures—regime shifts can be triggered by strong, short-lived exogenous forcing, but also by weaker, protracted pressures. These types of regime shifts can be hard to identify, because they lack the diagnostic “abrupt shift” from one state to another. We present evidence from simulation models and long-term eutrophication, grazing, and fire press experiments in grasslands. The results suggest that following a rapid change in exogenous forcing, slower-moving ecosystems can operate beyond a magnitude-threshold (a bifurcation point) for 3-20 years before ecosystem transitions become irreversible on human time scales. When the duration of exogenous forcing exceeds this “critical press time”, patches or entire ecosystems remain trapped in a new regime, despite reductions in exogenous pressures. However, some regime shifts can be averted if these pressures are reduced before they exceed their critical press time. On the other hand, inertia may overwhelm adaptive efforts in certain situations. For example, rapid nutrient and pollutant addition may have a critical press time of just 1 to 2 years. Considering the widespread nature of global change presses and slower rate of transition in many terrestrial ecosystems, the concept of critical press time highlights a buffering capacity that could be leveraged to navigate undesirable thresholds and prioritize ecosystem restoration. LTER data and experiments are uniquely poised to study the temporal dimension(s) of critical transitions. To this end, we provide methods for identifying the existence of critical thresholds in press experiments and critical press times.