Reliable hunting bag statistics are central for informed wildlife management. In the absence of complete reporting, hunting harvest must be estimated based on partial data, which requires reliable data and appropriate statistical methods. In the Swedish system, hunting teams, whose positions are known to the level of Hunting Management Precincts (HMPs), report their harvest of open season game and the size of the land on which they hunt, and the harvest on the non-reported area is estimated based on the reports. In this study, we improved data quality by solving several identified issues in the spatial data and provided temporally consistent estimates of huntable land (EHL) based on documented assumptions. We applied a recently developed method, the Bayesian Hierarchical and Autoregressive Estimation of Hunting Harvest (BaHAREHH), to harvest reports of 34 species from 2003--2021, using both previous and updated EHL, and compared harvest estimates to previously available estimates using naïve linear extrapolation (LE), which has been used as Sweden’s official harvest statistics. We found that updating EHL had a minor effect on harvest estimates at the national level but sometimes had a large impact at the level of individual HMPs. At the national level, previous LE estimates were similar to updated BaHAREHH estimates for species harvested at large numbers, but discrepancies were observed for species harvested at low rates. Time series of harvest estimated with LE had exaggerated temporal trends, higher coefficient of variation, and lower autcorrelation. At the level of counties and HMPs, there were substantial differences for all species, with some harvest estimates differing by several orders of magnitude. We conclude that the previously available LE estimates are sensitive to individual reports that add variability to the estimates and are, for some species, unreliable, especially at the level of county and HMP.