A computerised system called 'Computerised records for in-patient medical patients with drug interaction screening' (CRIMP-DIS) was developed in BASIC programming language and implemented in the internal medicine section of St. Luke's general hospital. The length of hospitalisation was determined for each patient in an experimental ward and in a control ward. A cost-benefit analysis was carried out on the results obtained in determining the decrease in length of hospitalisation. The results obtained support the hypothesis that computerised medication records provide better information to pharmacists and this leads to better patient care and health, which is reflected by a decrease in the length of hospitalisation. This decrease renders the system economically feasible and accentuates the need for local computerised medication record systems.