Airside service roads are essential for the movement of ground support vehicles that enable efficient aircraft turnaround and overall airport operations. Despite their importance, no standardised method exists for evaluating their capacity and performance, limiting the ability of airports to optimise this operational infrastructure. However, the heterogeneity of service road traffic — characterised by vehicles with varying speeds, sizes, and manoeuvrability — complicates capacity estimation and Level of Service (LoS) assessment. This research shows that a combination of microscopic traffic simulation and Multiple Linear Regression (MLR) provides a robust framework for estimating Passenger Car Unit (PCU) values and evaluating service road performance. Using five common airside road configurations, the method estimates PCU values for eleven service vehicle types and establishes LoS boundaries expressed in PCU per hour. The results reveal that PCU values vary significantly across intersection types, with four-way priority intersections exhibiting the highest sensitivity to traffic heterogeneity. Roundabouts are found to facilitate smoother traffic flow. Two three-way intersection types provide similar PCU values, but traditional priority rules facilitate substantially better performance. The results of the component-level models are upscaled to a service road network model, presenting a method for describing network performance. These findings provide a standardised method for assessing airside service roads, enabling airports to improve infrastructure planning and mitigate congestion amid rising aviation demands.