This study presents pseudo-proxy experiments to quantify the reconstruction skill of two climate field reconstruction methodologies for a marine proxy network subject to age uncertainties. The BARCAST methodology (Bayesian Algorithm for Reconstructing Climate Anomalies in Space and Time) is tested for sea surface temperature (SST) reconstruction for the first time over the northern North Atlantic region, and compared with a classic analogue reconstruction methodology. The reconstruction experiments are performed at annual and decadal resolution. We implement chronological uncertainties inherent to marine proxies as a novelty, using a simulated age-model ensemble covering the past millennium. Our experiments comprise different scenarios for the input data network, with the noise levels added to the target variable extending from ideal to realistic. Results show that both methodologies are able to reconstruct the Summer mean SST skillfully when the proxy network is considered absolutely dated, but the skill of the analogue method is superior to BARCAST. Only the analogue method provides skillful correlations with the true target variable in the case of a realistic noisy and age-uncertain proxy network. The spatiotemporal properties of the input target data are partly contrasting with the BARCAST model formulations, resulting in an inferior reconstruction ensemble that is similar to a white-noise stochastic process in time. The analogue method is also successful in reconstructing decadal temperatures, while BARCAST fails. The results contribute to constraining uncertainties in CFR for ocean dynamics which are highly important for climate across the globe.