Objective: The development of a new non-intrusive optical system for remotely measuring acoustic startle reflex (ASR) in humans. Approach: The eye reflex movement during an acoustic stimulation session is recorded through a high-speed digital camera. The eyes region is isolated by the rest of the face by an advanced pyramid-like feature detection algorithm, which greatly reduces the number of false positives. A separate Lucas–Kanade optical flow routine is designed for the eyeblink movement detection and the startle eyeblink reflex (SEBR) curve extraction. Image masking is implemented for the elimination of unwanted artifacts caused mainly by voluntary eye movement. The proposed system was tested along with a valid EMG system on a sample of 32 healthy randomly selected adults, and the results were compared in order to measure the system’s degree of reliability. Main results: To assess the proposed method’s validity the EMG data was used as a benchmark. The results showed strong correlation between EMG and Camera acquired results, which proves the validity of the proposed method. Furthermore, by comparing the response probability and the signal to noise ratio (SNR) for the two techniques, we proved that the proposed method can surpass the traditional EMG system in terms of accuracy and reliability. Significance: The proposed technique presents a simple, robust and reliable non-intrusive means of measuring ASR in humans, with the potential of future implementation on various ASR psychophysiology experiments, such as the study of PPI.