Flexible pavements are essential transportation assets, and their deterioration under mixed traffic conditions creates significant maintenance, safety, and economic challenges. Although traffic loading is widely recognized as being associated with pavement performance, relatively few studies have developed locally calibrated and easily interpretable regression models that quantify the contribution of different vehicle classes to pavement condition at the corridor level. This study applies multiple linear regression (MLR) to analyze the relationship between vehicle composition and the Pavement Condition Index (PCI) along a 3.5 km road section in Purwakarta, Indonesia, consisting of 35 pavement segments. Primary data were collected through pavement distress surveys and traffic observations categorized into light, medium, and heavy vehicles. The surveyed corridor exhibited a mean PCI of 61.89, indicating generally good pavement condition, although variability in distress severity was observed, with rutting and potholes as dominant types. The MLR model explained 79.4% of PCI variability ($R^2$ = 0.794, $p <$ 0.001). Heavy vehicles exhibited the strongest statistically significant negative association with PCI, followed by medium vehicles, whereas light vehicles were not statistically significant after adjustment for the remaining traffic categories. These findings indicate that traffic composition, particularly heavy-vehicle traffic, provides valuable information for pavement maintenance planning.