People’s estimates are biased toward previously considered numbers (anchoring). We have aggregated all available data from anchoring studies that included at least two anchors into one large dataset. Data were standardized to comprise one estimate per row, coded according to a wide range of variables, and are available for download and analyses online (https://metaanalyses.shinyapps.io/OpAQ/). Because the dataset includes both original and meta-data it allows for fine-grained analyses (e.g., correlations of estimates for different tasks) but also for meta-analyses (e.g., effect sizes for anchoring effects).