This study aims to optimize the efficiency of a high-traffic fuel station in Mumbai, India, by employing queuing theory and analyzing the service time distribution. Data was collected from a station with four fuel nozzles during rush hours, including inter-arrival times, service times, waiting times, fuel types, and vehicle types. Inter-arrival time data followed a Poisson distribution, while the service time data for each nozzle was analyzed using various distributions. The analysis revealed that the Weibull distribution provided the best fit for the service time data. This study utilized a multi-server queueing model (M/G/4) to calculate system utilization, waiting times, and the mean number of customers in the queue. According to the result, the model output was reasonably matched to observed data as the error percentage was found less than 20% for all output parameters. Several recommendations were made to optimize the fuel station, including automation, analysis, and maintenance. Practical implementation of low-budget modification and non-technical recommendations reduced the system’s waiting time by 22%. For long-term optimization, further data analysis and implementation of the recommendations are necessary. Nevertheless, this study provided valuable insights for fuel station operators in India.