Spectroscopic observations using the Infrared Space Observatory are reported towards the two well known infrared sources and young stellar objects L1551 IRS 5 and L1551 NE, and at a number of locations in the molecular outflow. The ISO spectrum contains several weak gas-phase lines of O I, C II, ?Fe II? and ?Si II?, along with solid state absorption lines of CO, COensuremath 2ensuremath<?subensuremath>, Hensuremath 2ensuremath<?subensuremath>O, CHensuremath 4ensuremath<?subensuremath> and CHensuremath 3ensuremath<?subensuremath>OH. Hubble Space Telescope (HST) images with the NICMOS infrared camera reveal a diffuse conical shaped nebulosity, due to scattered light from the central object, with a jet emanating from L1551 IRS 5. The continuum spectral energy distribution has been modelled using a 2D radiative transfer model, and fitted for a central source luminosity of 45 L$ødot$, surrounding a dense torus extending to a distance of textttchar126 3x104 AU, which has a total mass of textttchar126 13 M$ødot$. The visual extinction along the outflow is estimated to be $approx$ 10 and the mid-plane optical depth to L1551 IRS 5 to be $approx$ 120. This model provides a good fit to the ISO spectral data, as well as to the spatial structures visible on archival HST?NICMOS data, mid-IR maps and sub-millimetre radio interferometry, and to ground-based photometry obtained with a range of different aperture sizes. On the basis of the above model, the extinction curve shows that emission at wavelengths shorter than textttchar126 2 ensuremathmum is due to scattered light from close to L1551 IRS 5, while at wavelengths $gtrsim$ 4 ensuremathmum, is seen through the full extinguishing column towards the central source. Several ?Fe II? lines were detected in the SWS spectrum towards L1551 IRS 5. Although it would seem at first sight that shocks would be the most likely source of excitation for the ?Fe II? in a known shocked region such as this, the line intensities do not fit the predictions of simple shock models. An alternative explanation has been examined where the ?Fe II? gas is excited in hot ( textttchar126 4000 K) and dense ($gtrsim$ 10ensuremath 9ensuremath<?supensuremath> cmensuremath -3ensuremath<?supensuremath>) material located close to the root of the outflow. The SWS observations did not detect any emission from rotationally excited Hensuremath 2ensuremath<?subensuremath>. Observations with United Kingdom Infrared Telescope (UKIRT) of the vibrationally excited S- and Q-branch lines were however consistent with the gas having an excitation temperature of textttchar126 2500 K. There was no evidence of lower temperature ( textttchar126 500 K) Hensuremath 2ensuremath<?subensuremath> gas which might be visible in the rotational lines. Observations with UKIRT of the CO absorption bands close to 2.4 ensuremathmum are best fit with gas temperatures textttchar126 2500 K, and a column density textttchar126 6x10ensuremath 20ensuremath<?supensuremath> cmensuremath -2ensuremath<?supensuremath>. There is strong circumstantial evidence for the presence of dense (coronal and higher densities) and hot gas (at least 2500 K up to perhaps 5000 K) close to the protostar. However there is no obvious physical interpretation fitting all the data which can explain this.