Context. The Gaia-ESO Survey (GES) is a large public spectroscopic survey that has collected, over a period of six years, spectra of ∼ 10 5 stars. This survey provides not only the reduced spectra, but also the stellar parameters and abundances resulting from the analysis of the spectra. Aims. The GES dataflow is organised in 19 working groups. Working group 13 (WG13) is responsible for the spectral analysis of the hottest stars (O, B, and A type, with a formal cutoff of T eff > 7000 K) that were observed as part of GES. We present the procedures and techniques that have been applied to the reduced spectra in order to determine the stellar parameters and abundances of these stars. Methods. The procedure used was similar to that of other working groups in GES. A number of groups (called Nodes) each independently analyse the spectra via state-of-the-art techniques and codes. Specific for the analysis in WG13 was the large temperature range covered (T eff ≈ 7000 − 50 000 K), requiring the use of different analysis codes. Most Nodes could therefore only handle part of the data. Quality checks were applied to the results of these Nodes by comparing them to benchmark stars, and by comparing them to one another. For each star the Node values were then homogenised into a single result: the recommended parameters and abundances. Results. Eight Nodes each analysed part of the data. In total 17 693 spectra of 6462 stars were analysed, most of them in 37 open star clusters. The homogenisation led to stellar parameters for 5584 stars. Abundances were determined for a more limited number of stars. The elements studied are He, C, N, O, Ne, Mg, Al, Si, and Sc. Abundances for at least one of these elements were determined for 292 stars. Conclusions. The hot-star data analysed here, as well as the GES data in general, will be of considerable use in future studies of stellar evolution and open clusters.