We demonstrate the feasibility of growing crystals of protein in volumes as small as 1 nanoliter, Advances in the handling of very small volumes (i.e. through ink-jet and other technologies) open the way towards fully automated system,,. The rationale for these experiments is the desire to develop a system that speeds up the structure determination of proteins by crystallographic techniques, where most of the precious protein sample is wasted for the identification of the ideal crystallisation conditions, An additional potential benefit of crystallisation in very small volumes is the potential improvement of the crystal quality through reduced convection during crystal growth. Furthermore, in such small volumes even very highly supersaturated conditions can be stable for prolonged periods, allowing additional regions of phase-space to be prospected for elusive crystallisation conditions. A massive improvement in the efficiency of protein crystallogenesis will cause a paradigm shift in the biomolecular sciences and will have a major impact in product development in (for example) the pharmaceutical industry.