This paper discusses flight test results of a precise relative navigation and separation assurance system based on the exchange of Global Navigation Satellite (GNSS) measurements via new message types for the Automatic Dependent Surveillance – Light (ADB-L) instead of Broadcast (ADS-B). With an increased use of Unmanned Aircraft Systems (UAS) for a wide variety of applications, such as law enforcement, search and rescue, agriculture, infrastructure inspection, maintenance, mapping, and journalism, it is expected that UAS will be taking off, landing, or otherwise operating at airfields at the same time as manned aircraft. These operations may lead to smaller separation between the participants and, thus, higher collision risk. To enable smaller separation, relative position and velocity estimators are required that can meet stricter navigation performance. Typically, ADS-B transmits traffic position and velocity estimates output by the onboard GNSS receiver. A measurement-based ADS-B implementation, which transmits raw measurements from the GNSS receiver rather than aircraft state vectors and performance parameters, has been proposed in previous papers by the authors to improve surveillance performance and add integrity to the surveillance solution. Test results of the proposed method have shown meter-level relative position accuracy and millimeter-per-second-level relative velocity accuracy. This paper will review these methods and their performance, propose an implementation on ADS-L, discuss how this level of performance can enable the simultaneous operation of manned and unmanned aircraft at low altitudes in the vicinity of airfields, and show how off-nominal operations can be detected (e.g., leaving the geofence or deviations from approved trajectories). The paper will furthermore discuss recent flight tests with one manned aircraft and two UAS as part of project Safefly and illustrate the benefits of using GNSS measurement-based ADS-L data for separation assurance as opposed to traditional methods.