e.. it includes the setup time (moving the patient on the table, injecting or feeding the patient with contrast, applying anaesthesia, etc.) as well as the scanning time was found that the total number of patients (including walk ins) investigated daily varies over the days of the week. The num- ber of outpatients (including walk-ins) investigated varies from around 55% to 75% (65% on an average) of the total number of patients investigated in a day, and that of inpatients varies from around 25% to 45% (35% on an average) of the total number of patients investigated. The data on arrival patterns revealed that there is unpunctuality and no-shows among the scheduled out patients. The punctuality level is different for outpatients and in patients. The mean service times are also different for inpatients and outpatients (see Table 1 for details). The inter-arrival times of walk-ins were assumed to follow an exponential distribution with a mean of 60 min. Walk-ins were assigned a lower priority than the scheduled arrivals. Average service times and service time variabil ity vary widely among the different types of investigations. was found that among all types of investigations, there is highest vari ability among service times for MRI of brain, screening studies and MR Angiogram studies. The data for arrival punctuality and service times were analysed using Arena input analyser in order to obtain expressions for the distributions followed by them. It was found that the empirical form ofarrival punctuality data did not fit to any particular theoretical distribution. However, it was observed that the outpatients were much more unpunctual as compared to inpa- tient arrivals. This is obvious as the inpatients are already staying in the hospital and their arrivals are more under control. For our anal yses, we assumed that the punctuality data followed a uniform dis tribution. The punctuality of scheduled outpatients was assumed to follow a Uniform (-15, 20) distribution and that of inpatients was assumed to follow a Uniform (-10, 10) distribution. was found that the average service times for outpatients were shorter than inpatients. This is because the inpatients usually may not be able to move themselves and require support from staff. With the arrival punctuality distribution known, we can derive expressions for inter-arrival time distributions. Inter-arrival times for outpa- tients and inpatients in this case followed a Triangular distribution 132]. The mean service time of all the patients is 39.3 min and the standard deviation o is 13,6 min