In recent years, the Internet has become a popular source for gathering and collecting information, especially in online collaborative projects such as Wikipedia or OpenStreetMap. In these projects, collaboration resembles a sequential chain that starts with the creation of an entry followed by a sequence of contributors deciding to adjust or maintain the presented information. As online collaborative projects were found to yield highly accurate information which is often attributed to wisdom of crowds, we examine this sequential collaboration as a process of judgment aggregation. Thereby, sequential collaboration resembles advice taking since contributors encounter judgments of previous participants before deciding whether to adjust or maintain these judgments. In three experiments, comparing judgment aggregation with sequential collaboration and the unweighted averaging of independent individual judgments, we found that judgment accuracy in sequential collaboration increases over a sequential chain and that sequential-collaboration estimates can be more accurate than estimates obtained with unweighted averaging. By allowing contributors to opt-out of providing a judgment, sequential collaboration may foster an implicit weighting of judgments by expertise such that contributors adjust or maintain judgments according to their expertise. We investigated this in three experiments measuring and manipulating contributors' expertise. There we showed that experts improve judgments more than novices resulting in more accurate estimate the more and later experts enter sequential chains. These results yield first insights into sequential collaboration as a mechanism of judgment aggregation and show that advice taking in the context of sequential collaboration works to the benefit of the resulting judgments.