We investigate the feasibility of deep-space navigation using the highly stable periodic signals from X-ray pulsars in combination with dedicated instrumentation on the spacecraft: a technique often referred to as 'XNAV'. The results presented are based on the outputs from a study undertaken for the European Space Agency. The potential advantages of this technique include increased spacecraft autonomy and lower mission operating costs. Estimations of navigation uncertainties have been obtained using simulations of different pulsar combinations and navigation strategies. We find that the pulsar PSR B1937 + 21 has potential to allow spacecraft positioning uncertainties of~2 and~5 km in the direction of the pulsar after observation times of 10 and 1 h respectively, for ranges up to 30 AU. This could be achieved autonomously on the spacecraft using a focussing X-ray instrument of effective area~50 cm 2 together with a high performance atomic clock. The Mercury Imaging X-ray Spectrometer (MIXS) instrument, due to be launched on the ESA/JAXA BepiColombo mission to Mercury in 2018, is an example of an instrument that may be Exp Astron (2016) further developed as a practical telescope for XNAV. For a manned mission to Mars, where an XNAV system could provide valuable redundancy, observations of the three pulsars PSR B1937 + 21, B1821-24 and J0437-4715 would enable a three-dimensional positioning uncertainty of~30 km for up to 3 months without the need to contact Earth-based systems. A lower uncertainty may be achieved, for example, by use of extended observations or, if feasible, by use of a larger instrument. X-ray instrumentation suitable for use in an operational XNAV subsystem must be designed to require only modest resources, especially in terms of size, mass and power. A system with a focussing optic is required in order to reduce the sky and particle background against which the source must be measured. We examine possible options for future developments in terms of simpler, lower-cost Kirkpatrick-Baez optics. We also discuss the principal design and development challenges that must be addressed in order to realise an operational XNAV system.