Using a comprehensive simulation model, a number of important parameters relating to manufacturing cells are investigated. These include: (i) the proposal by Burbidge that ‘period batch control’ is more appropriate for group technology cells, (ii) the effects of different numbers of tool set-up types or part families, (iii) the comparative merits of a larger 18-machine cell and 3 parallel 6-machine cells and (iv) the effects of inter-operation transit time within cells. System performance obtained showed that, with an efficient scheduling rule, the smaller batches product better facility utilization and job throughput. The results also suggest that, expect for extremely small, restrictive cells (with one or two tool setup types), the addition of further part families does not cause a serious deterioration in the productivity of the cell. The larger cells were found to be more effective than the small cells placed in parallel in all the system performance characteristics recorded.