Whole-tree harvest (WTH) may compromise tree productivity and health and lead to soil and surface water acidification. The aim of this study was to assess the long-term change (up to circa 40 years) in soil exchangeable calcium (Ca2+) pools following conventional (CH) and WTH at three Swedish coniferous sites. A second aim was to evaluate how well the results could be reproduced by the dynamic model MAGIC (Model of Acidification of Groundwater in Catchments). Soil Ca2+ pools (down to 20 cm) decreased at all three sites from stand age 15–16 to 37–38 years. The depletion ranged from 2.6 to 8.6 kEq ha−1 (26.5–52.7%) and 0.2 to 5.0 kEq ha−1 (2.3–49.1%) in the CH and WTH treatment, respectively. The presence of an interaction effect indicated that the main effect of time was not statistically significant at all three sites. Over the course of time, soil Ca2+ pools have also become more similar between the CH and WTH treatments, but the Ca2+ pools were still significantly lower (p < 0.05) after WTH at stand age 37–38 years. The measured declines in Ca2+ pools were generally greater than what has been found in other studies and were largely explained by high soil Ca2+ availability and high tree Ca2+ uptake, especially in the CH-plots, as indicated by the MAGIC mass balance budgets. Model simulations by MAGIC partly agreed with the measured data. However, the model exaggerated the soil Ca2+ losses between 1990 and 2013 (CH = 3.6–9.9 kEq ha−1; WTH = 3.0–8.3 kEq ha−1), especially at the spruce sites. Furthermore, MAGIC could not reproduce the rapid diminishing differences between CH and WTH. Uncertainties in model parameters, underestimated soil Ca2+ pools or biological feed-back mechanisms could explain this discrepancy. Until these have been resolved, interpretations of Ca2+ changes related to CH or WTH using dynamic modelling or mass balance budget calculations should be done with caution