A full account is given of the relatively new astronomical technique of multi- object spectroscopy with optical fibre lightguides. The thesis begins with a survey of the number densities of a wide range of astronomi- cal object classes, and, after comparing the merits of the various techniques for multi- object spectroscopy, shows (by defining the "effective aperture" of a multiple fibre - coupled telescope) that the multi -fibre tech- nique is very well- suited to the real distribution of celestial objects. A review of the properties of optical fibres is given, together with an exposition of the important considerations in the design of fibre feeds for astronomy. This is followed by a description of the requirements in the design of telescopes, spectrographs and detectors to render them suitable for the multi -fibre technique. Equations are derived for the performance of these systems in terms of signal -to -noise ratio, and an example is given. A survey of the equipment that has been built for multi -fibre spectroscopy at observatories throughout the world is presented, and special attention is paid to the FOCAP fibre system at the 3.9 -metre Anglo- Australian tele- scope. The thesis gives an example of the use of FOCAP in work carried out by the writer to investigate the kinematics of RR Lyrae variables in the nuclear bulge of the Galaxy. Evidence is found in the preliminary results for the existence of non -circular motions along the galactic centre line -of- sight. A description is then given of the FLAIR wide -field fibre -coupled multi- object spectroscopy system built by the writer for the 1.2 -metre UK Schmidt Telescope. Reports are presented on a series of observational pro- grammes designed to test the system, including the classification of stars in the Chamaeleon dark cloud, spectroscopy of Orion cluster flare stars, spectroscopy of objects in the Large Magellanic Cloud, the measurement of galaxy redshifts, and the observation of quasar candidates. An experiment in time -resolved multi- object photometry is also described. It is shown that for spectroscopy, the prototype system will perform well on stellar targets to V "16 -17, but, although galaxy redshifts have been obtained, its usefulness for the systematic measurement of redshifts is hampered by its low sensitivity in the blue region of the spectrum. An account is then given of the relatively simple steps needed to rectify this, together with a brief description of a proposed, fully- optimised FLAIR system. Finally, the future prospects for multi -fibre astronomical spectroscopy are briefly reviewed, from the point of view of the new instrumentation becoming available, and the astronomical results that will be produced.