Reconstructions of past climate behavior often describe prominent anomalous periods that are not necessarily captured in climate simulations. Here, we illustrate the contrast between an interdecadal strong positive phase of the winter Pacific/North American pattern (PNA) in the early 19th century that is described by a PNA reconstruction based on tree rings from northwestern North America, and a slight tendency towards negative winter PNA anomalies during the same period in an ensemble of state-of-the-art coupled climate simulations. Additionally, a pseudo-proxy investigation with the same simulation ensemble allows for assessing the robustness of PNA reconstructions using solely geophysi-cal predictors from northwestern North America for the last millennium. The reconstructed early 19th-century positive PNA anomaly emerges as a potentially reliable feature, although the pseudo-reconstructions are subject to a number of sources of uncertainty and deficiencies highlighted especially at multidecadal and centennial timescales. The pseudo-reconstructions demonstrate that the early 19th-century discrepancy between reconstructed and simulated PNA does not stem from the reconstruction process. Instead, reconstructed and simulated features of the early 19th-century PNA can be reconciled by interpreting the reconstructed evolution during this time as an expression of internal climate variability, which is unlikely to be reproduced in its exact temporal occurrence by a small ensemble of climate simulations. However , firm attribution of the reconstructed PNA anomaly is hampered by known limitations and deficiencies of coupled climate models and uncertainties in the early 19th-century external forcing and background climate state.< Réduire