Anchoring is one of the most studied and robust behavioral biases, but there is little knowledge about its persistence and magnitude in economic settings. This article studies the role of anchoring bias in strategic interactions. We test experimentally two different anchor types using auctions. The announcement of a random group identification number but also of an upper bid limit, despite being irrelevant for a rational player, result in higher bids in first-price sealed-bid auctions. In Dutch auctions, the effect of the bid limit is negative. Based on these results we demonstrate that the long-established ranking that the Dutch auction generates lower revenue than the first-price sealed-bid auction crucially depends on the size of the anchor. The observed adjustment behavior can be explained as a rational response to anchored biased beliefs.