We undertake an optical and ultraviolet spectroscopic analysis of a sample of 20 Galactic B0 - B5 supergiants of luminosity classes Ia, Ib, Iab and II. Fundamental stellar parameters are obtained from optical diagnostics and a critical comparison of the model predictions to observed UV spectral features is made. These parameters are derived for individual stars using CMFGEN, the nLTE, line-blanketed model atmosphere code of Hillier et al., 1998. The B supergiant temperature scale derived here shows a reduction of 1000 - 3000 K compared to previous results obtained using unblanketed codes. Mass loss rate estimates are in good agreement with predicted theoretical values and all of the 20 B0 - B5 supergiants analysed show evidence for CNO processing. The observed WLR values calculated for B0 - B0.7 supergiants are larger than predicted values, whereas the reverse is true for B1 - B5 supergiants. This means that the discrepancy between observed and theoretical values cannot be resolved by adopting clumped (i.e., lower) mass loss rates, as for O stars. The most surprising result is that, although CMFGEN succeeds in reproducing the optical spectrum accurately, it fails to reproduce key UV diagnostics, such as NV and CIV P Cygni profiles, precisely. This problem arises because the models are not ionised enough and fail to reproduce the full extent of the observed absorption trough of the P Cygni profiles. These findings add further support to the need to revise the standard model of massive star winds.

We present empirically determined ionization conditions for the winds of 106 luminous B0 to B5 stars observed by IUE. The UV wind lines are modelled to extract products of mass-loss rates (Mdot) times ionization fractions (Mdot q_i(w), where w = v/v_infty) for N V, C IV, Si IV, Si III, Al III and C II. We describe the general behaviour of the Mdot q_i(w) and their ratios, demonstrating that the wind ionization increases with distance from the star, contrary to recent findings for O star winds. Using empirical mass-loss rates (from H alpha observations) and model prescriptions, we derive mean q_i(w) values integrated over the wind, <q_i>. These <q_i> are quite small, never exceeding 15% for Al III or 2% for Si IV. This is surprising, since the <q_i> for these ions clearly peak within the observed spectral range. We conclude that the low <q_i> arise because the <Mdot q_i> are underestimated by the wind models, which assume that the outflows are smooth when they are, in fact, highly structured.