A rear-end collision avoidance system (RECAS) could reduce accidents by assisting drivers during emergencies through autonomous braking. If the decision to intervene is made too early, the intervention can become a disturbance to the driver and if the decision is made too late, the safety benefits of the system will be reduced. Hence, proper threat assessment is important to make a decision for intervention in a RECAS. In this study, a new threat assessment measure using a varying time headway has been proposed to develop a rear-end collision avoidance algorithm (RECAA) for heavy commercial road vehicles in heterogeneous traffic. The proposed varying time headway depends on the type of lead vehicle, host vehicle load conditions, road conditions and relative longitudinal speed. The developed RECAA was implemented in a hardware-in-loop experimental setup that is equipped with the vehicle dynamic simulation software, IPG/TruckMaker® and tested for vehicle following scenarios using different types of lead vehicles and host vehicle loading conditions. The results showed that the proposed varying time headway improved the RECAA by activating the host vehicle's brake system at different time instants based on the type of lead vehicle and thereby preventing the unintended early brake intervention of RECAS. © 2017 Elsevier Ltd