Although expertise strongly influences the accuracy of judgments, its role in anchoring is still unclear with findings of reduced, similar, and stronger anchoring in experts compared to novices. Similarly, the three most prominent theories of anchoring – the Insufficient Adjustment Model, the Selective Accessibility Model, and the Scale Distortion Theory – also make different predictions regarding the influence of expertise on anchoring. To address this inconsistency in both empirical results and theoretical accounts, we manipulate individuals' expertise prior to a perceptual estimation task. Based on the literature on numerical judgments and wisdom of crowds, we expect participants trained to become experts are less influences by anchors than untrained participants. Additionally, we manipulate anchor relevance and anchor extremity for which all three theories make similar prediction. Here, we expect stronger anchoring effects for extreme compared to moderate anchors and relevant compared to irrelevant anchors. In two preregistered experiments (total N = 567), we find that experts do indeed show less anchoring compared to novices, and that more extreme anchors lead to stronger anchoring effects. However, the effect of anchor relevance is inconclusive in both experiments. In a third upcoming study, we will manipulate participants expertise before a cued judgment tasked. Again, we expect that anchoring is reduced for experts, irrelevant anchors, and moderate anchors. Our results demonstrate that expertise has considerable effects on anchoring. Therefore, theories of anchoring should take expertise into account as a strong inhibitor of anchoring effects.