Fear conditioning (FC) allows the dissociation of hippocampal and nonhippocampal behavioral function in rodents, and has become a diagnostic tool in transgenic mouse research employed to investigate mutation-induced changes in brain function. Although the procedural details of the paradigm have been established, quantification of the behavioral output, freezing, remains problematic in mice. Observation-based techniques are time-consuming and may be subject to bias, while movement detection with photocells is imprecise. Here we describe an alternative method for movement detection, based on an electronic force transducer system that allows the quantification of acceleration forces generated by a moving subject. We compare the behavior of two inbred strains of mice (C57BL/6 and DBA/2) whose performance is known to differ in hippocampal tasks, including FC. The comparison is made using multiple techniques: the force transducer approach, and three observation-based methods, a computer-aided event-recording approach, a traditional time-sampling paper/pencil method, and a subjective impression-based scoring system. In addition, we investigate the correlation structures of behavioral elements quantified by event recording, using principal component analyses; we conclude that fear may manifest in multiple forms and in a stimulus- and genotype-dependent manner. We suggest that the force transducer system provides precise quantification of movements in an automated manner and will allow high-throughput screening for mutation and drug effects in mice. However, we also argue that fear responses can be complex, and freezing behavior may not be the only measure of fear or fear-associated memory.