Transit agencies are increasingly partnering with ridehailing platforms to address first- and last-mile access barriers, but empirical evidence on how riders use Transportation Network Companies (TNCs) within operational Mobility-as-a-Service (MaaS) systems remains limited. This study examines Uber-transit linkages in the Vamos-EZHub MaaS platform in San Joaquin County, California, where Uber was integrated alongside transit ticket activations. Using platform telemetry, Uber trip records, transit ticket activations, GTFS stop and service data, and stop-area built-environment measures, we identify Uber trips that plausibly connect to transit using an asymmetric spatiotemporal rule. First-mile links are defined as Uber dropoffs within 250 meters of a transit stop followed by a transit activation within 30 minutes; last-mile links are defined as Uber pickups within 250 meters of a stop occurring within 90 minutes after a transit activation.

Results show that Uber-transit linkages represent a small but behaviorally meaningful subset of platform activity. We identify 167 linked Uber trips, including 119 last-mile trips, 32 first-mile trips, and 16 ambiguous links. Linked trips are spatially concentrated at a limited set of stops and temporally patterned by time of day and day of week. Stop-level screening models suggest that linked trips are associated with characteristics of the transfer-stop environment, particularly lower car access and selected land-use contexts, rather than being evenly distributed across the network.

These findings provide real-world evidence on how ridehailing is used as a transit access mode within a public MaaS platform. While linked Uber-transit use remains modest, its concentration around specific stops and neighborhoods suggests that TNC integration may serve targeted access needs rather than functioning as a broad substitute for transit.