The Boreal Avian Modelling (BAM) Project uses statistical offsets to correct for methodology and detectability differences across species, projects, and surveys times/locations as described in Solymos et al.. We refer to this approach as the QPAD approach. QPAD version 1 was an internal working version of the parameter estimates. QPAD version 2 was published alongside Solymos et al. and its supporting information. Since closing the data for this manuscript, more data have accumulated from Canada and the United States as part of BAM’s point count database. QPAD version 3 represents a major update of the parameter estimates. The methodology is the same as outlined in the paper, but the data set is different in terms of number of surveys, species that met sample size requirements, and some of the predictor variables as outlined in the following sections. This technical report documents and explains the QPAD version 3 estimates.