We describe a real-time automatic target recognition (ATR) system that employs a state-of-the-art optical correlator. The system’s ability to acquire and track a moving vehicle target against a densely cluttered natural background is investigated using six test image sequences obtained from a scanning IR sensor array mounted on an airborne platform. The approach adopted for the evaluation is to first optimize a number of system parameters, using what we considered to be one of the better test sequences, and then to test the system with the optimum configuration on the remaining sequences. The system performance is quantified by measuring its receiver-operating-characteristic curves against each test image sequence. Although successful acquisition and tracking of the target is demonstrated for some test sequences, there are numerous occasions on which the system acquired and tracked false alarms. This is primarily because edge features alone are an insufficient discriminator. We therefore conclude that the system evaluated does not exhibit the desired robustness for the acquisition and tracking of vehicles in densely cluttered natural scenes. Finally, these results are compared with those produced using a digital simulation of the ATR system’s algorithmic processing chain. Comparison shows that for all practical purposes the performances are equivalent. © 1999 Society of Photo-Optical Instrumentation Engineers.