A central question in the study of LTP has been to determine what role it plays in memory formation and storage. One valuable form of learning for addressing this issue is associative fear conditioning. In this paradigm an animal learns to associate a tone and shock, such that subsequent presentation of a tone evokes a fear response (freezing behavior). Recent studies indicate that overlapping cellular processes underlie fear conditioning and LTP. The fear response has generally been scored manually which is both labor-intensive and subject to potential artifacts such as inconsistent or biased results. Here we describe a simple automated method that provides unbiased and rapid analysis of animal motion. We show that measured motion, in units termed significant motion pixels (SMPs), is both linear and robust over a wide range of animal speeds and detection thresholds and scores freezing in a quantitatively similar manner to trained human observers. By comparing the frequency distribution of motion during baseline periods and to the response to fox urine (which causes unconditioned fear), we suggest that freezing and non-freezing are distinct behaviors. Finally, we show how this algorithm can be applied to a fear conditioning paradigm yielding information on long and short-term associative memory as well as habituation. This automated analysis of fear conditioning will permit a more rapid and accurate assessment of the role of LTP in memory.