The paper by F. van Leeuwen evaluates the new Hipparcos astrometric catalogue, which claims to improve the accuracy of parallaxes for stars brighter than magnitude Hp = 8 by up to a factor of 4 compared to the original 1997 catalogue. The study checks the quality of the data, the consistency of formal errors, and the presence of error correlations. The internal errors are traced through the reduction process, and external errors are investigated using radio observations of a subset of stars and the distribution of negative parallaxes. The results confirm the formal errors on the parallaxes and suggest a small amount of additional noise, though this is unlikely. The new reduction provides a 2.2-fold improvement in total weight compared to the 1997 catalogue, offering enhanced data for studies on stellar luminosities and local galactic kinematics. The paper also addresses systematic errors in the original catalogue, the impact of improved attitude reconstruction, and the significance of abscissa-residual error correlations. Overall, the new catalogue is shown to be more accurate and reliable, with reduced correlation levels and improved data for a wide range of astrophysical applications.The paper by F. van Leeuwen evaluates the new Hipparcos astrometric catalogue, which claims to improve the accuracy of parallaxes for stars brighter than magnitude Hp = 8 by up to a factor of 4 compared to the original 1997 catalogue. The study checks the quality of the data, the consistency of formal errors, and the presence of error correlations. The internal errors are traced through the reduction process, and external errors are investigated using radio observations of a subset of stars and the distribution of negative parallaxes. The results confirm the formal errors on the parallaxes and suggest a small amount of additional noise, though this is unlikely. The new reduction provides a 2.2-fold improvement in total weight compared to the 1997 catalogue, offering enhanced data for studies on stellar luminosities and local galactic kinematics. The paper also addresses systematic errors in the original catalogue, the impact of improved attitude reconstruction, and the significance of abscissa-residual error correlations. Overall, the new catalogue is shown to be more accurate and reliable, with reduced correlation levels and improved data for a wide range of astrophysical applications.