Purpose: Many researchers require access to computer facilities beyond those offered by desktop workstations. Traditionally, these are offered either through partnerships, to share the cost of supercomputing facilities, or through purpose-built cluster facilities. However, funds are not always available to satisfy either of these options, and university departments are under increasing pressure to obtain the maximum return on investment. The purpose of this paper is to describe the development of a low-cost cluster computing facility using existing workstations in undergraduate computing laboratories. Design/methodology/approach: Drawing on a previous feasibility study, experiments were conducted with cluster configurations of increasing size to determine the benefits and drawbacks. This was followed by identification and resolution of problems, and assessment of effects upon existing users of the computers. Findings: A working cluster was built to supply the needs of researchers, with almost no impact upon existing users and at very little cost. Research limitations/implications: At present, the workstations can only be used as a "processor farm" and it is unclear how MPI type processing may be achieved, because there is no guarantee of the number of workstations available at any specific time. Practical implications: Any institution requiring access to high performance computing can build a useful resource from existing computers at minimal cost. The usefulness of the resource depends on existing computers being under utilised. Originality/value: The novel aspect of this work is that the facility was built using existing student computing laboratories. (Contains 1 table and 3 figures.)