Subsidence curves, based on decompacted sediment thicknesses, are generally displayed without indicating depth or age errors. This makes their interpretation ambiguous, leading to incorrect identification of subsidence pulses and incorrect identification of periods of steady subsidence. Error analysis should also form part of thermal maturation studies based on such subsidence curves. Formal error analysis of subsidence curves is a complex problem because of the iterative and nonlinear nature of the decompaction calculations. This paper presents an alternative approach, based on Monte Carlo simulation, that allows subsidence curves to be calculated along with error estimates. We also show that identification of subsidence pulses and identification of periods of steady subsidence cannot be made on the basis of such subsidence curves, even when errors are shown, because errors on successive points on the subsidence curves are correlated. Instead, an ensemble of subsidence rate curves must be calculated at the same time as the ensemble of subsidence curves so that a mean subsidence rate curve can be calculated along with associated error estimates.