Introduction: Gap between demand and limited availability of resources often lead to delays, which in turn affect the patient flow in a typical hospital setting. Registration of new patients at front desk is one such place. Delays in the registration process can have an unwanted effect on patient-hospital relationship. Hence, it is important to reduce the delay. Aim: To identify various reasons for delay in the registration process of a multispecialty hospital, and develop a statistical model to estimate number of desks to be operated to reduce the wait time. Materials and methods: In a period of two weeks, a total of 923 new patient registrations were observed to collect the time consumed for various events of the registration process. Correlation analysis and statistical modelling are used in the study. Results and discussion: It was found that various factors such as literacy level of patients, mode of payments, age etc. do not show significant correlation with the registration time, except for the hourly patient load. The number of patients arriving in an hour (patient load) showed a significant correlation, suggesting that number of desks operated on regular basis was not sufficient to handle occasional high loads. A simple model that accounts for hourly patient load, estimating the number of desks to be operated in a given hour, to minimize the wait times at registration desk has been suggested. Conclusion: The model indicates that in the current setting, most of the times the resources were either underutilized or overloaded. The delays in registration process can be reduced by operating optimal number of desks as determined by the proposed model (i.e. function of hourly patient load).