["Legal and Criminological Psychology, Volume 26, Issue 2, Page 215-227, September 2021. ", "\n\nPurpose\nWhen making judgements under uncertainty not only lay people but also professional judges often rely on heuristics like a numerical anchor (e.g., a numerical sentencing demand) to generate a numerical response. As the prosecution has the privilege to present its demand first, some scholars have speculated about an anchoring‐based unfair disadvantage for the defence (who has the last albeit less effective word in court). Despite the plausibility of this reasoning, it is based on a hitherto untested assumption that the first of two sequential anchors exerts a greater influence on a later judgement (a primacy effect). We argue that it is also conceivable that the last word in court has a recency advantage (a recency effect) or that order does not matter as both demands even each other out (a combined anchor).\n\n\nMethods\nWe report a pre‐registered experiment with German law students (N = 475) who were randomly assigned to six experimental conditions in a study on legal decision‐making order to test these three possibilities.\n\n\nResults\nResults indicate an influence of both the prosecution and the defence recommendation, but no effect of order.\n\n\nConclusion\nThis provides strong support for combined anchoring even for knowledgeable participants and rich case material. Specifically, the data are best compatible with the notion that both anchors exert an influence but each on different individuals. The implications of this finding for theory and legal decision‐making are discussed.\n\n"]