We present the analysis and modeling of the emission spectra of the jovian northern auroral region taken from May 28 to June 3, 1993, with the Goddard High Resolution Spectrograph on board the Hubble Space Telescope. They extend from 1204 to 1240 Â covering H Lyman α and part of the Werner and the Lyman bands of H 2. We used the 2×2 arcsec large science aperture combined with the G160M grating (spectral resolution of 570 mÂ) centred on Jupiter's central meridian. The auroral region studied extends from 50 to 60° north latitudes and from 130 to 220° System III longitudes. Within the 1 arcsec pointing uncertainty, most of the region delineated by the theoretical ovals at 5.9 and 30 jovian radii (R J) in the VIP4 model of J. E. C. Connerney, M. H. Acuna, N. F. Ness, and T. Satoh (1998, New models of Jupiter's magnetic field constrained by the Io flux tube footprint. J. Geophys. Res. 103, 11,929–11,939), including the auroral oval derived from the Wide Field Planetary Camera 2 images by J. T. Clarke, G. E. Ballester, J. T. Trauger, R. Evans, J. E. P. Connerney, K. Stapelfeld, D. Crisp, P. D. Feldman, C. J. Burrows, S. Casertano, J. S. Gallagher, R. E. Griffiths, J. J. Hester, J. G. Hoessel, J. A. Holtzman, J. E. Krist, V. Meadows, J. R. Mould, P. A. Scowen, A. M. Watson, and J. A. Westphal (1996, Far-ultraviolet imaging of Jupiter's aurora and the Io “footprint”. Science 274, 404–409), was sampled. We derive auroral brightnesses (averaged over the slit) ranging from 20 kR in the 190–220° longitude region to 30 kR in the 130–160° longitude range. We use the theoretical model developed by D. Rego, R. Prangé, and L. Ben Jaffel (1999, Auroral Lyman α and H 2 bands from the giant planets. 3. Lyman α intensity and spectral profile including radiative effects and H 2 color ratios. J. Geophys. Res. Planets. 104, 5939–5954), which calculates self-consistently the auroral Lyman α line profile for electron and proton precipitations and a given frequency redistribution function. This model shows that Lyman α profiles are not dependent upon the identity of the particles for a given penetration depth. These profiles only constrain the atmospheric H column density above the emitting layer, H col. Best agreement with the data is found with the complete frequency redistribution (CR). We derive an auroral H col of 1.3×10 16 cm −2 for all spectra with an upper limit of 5×10 16 cm −2. Precipitating protons also produce fast H atoms by charge exchange which can also be excited, but the resulting Doppler-shifted profile was not detected. Present data thus rule out protons as being the only precipitating particles. In the auroral zones, theoretical models predict that chemical reactions induced by particle precipitations often result in the production of atomic hydrogen, and thus an enhancement of the H density is expected. However, the value of H col we derive, between ∼8 and 23 times less than the equatorial values of L. Ben Jaffel, J. T. Clarke, R. Prangé, R. Gladstone, and A. Vidal-Madjar (1993, The Lyman alpha bulge of Jupiter: Effects of a non-thermal velocity field. J. Geophys. Res. Lett. 20(8), 747–750), is in good agreement with the auroral values of R. Prangé, D. Rego, L. Pallier, L. Ben Jaffel, C. Emerich, J. T. Clarke, G. E. Ballester, and J. Ajello (1997, Detection of self-reversed Lyman α lines from the jovian aurorae with the Hubble space Telescope, Astrophys. J. 484, L169–L173). This suggests that the H abundance predicted with standard atmospheric model is unrealistically large in the auroral region.