The primary approach to measure hyper-redundant animal body structures is the use of high speed cameras in a laboratory environment, which can deprive locomotion of its proper context. Challenging conditions and complex three dimensional (e.g. rainforest or aquatic) environments make the collection of field data difficult, and prevents a complete analysis of an animal’s motion. We have developed liquid metal (eGaIn) based, hyper-elastic silicone strain sensors to measure local tail curvature with minimal impact on environment, mobility and body stiffness and therefore hope to enhance in situ biomechanics data collection without requiring manipulation of conditions. By not relying on imaging systems, long-duration data can be collected at very low latencies with minimal power and processing, and intricate movements can be measured in field experiments. This includes rapid tail surface righting, one of the first movement patterns observed in neonatal development. We propose utilizing soft sensors to measure subtle movements in aquatic animals as well as patterns of autotomized gecko tails. Ultimately, new insights into behavior, neuromuscular control and mechano-sensory receptivity can be gained. When connected to a soft undulating robotic fish with a tail beat frequency of 0.8-1.2 Hz, our sensor response is linear (R^{2(/sup) = 0.952) with a relative error that is well modeled by Gaussian noise (st. dev. of 0.4%). We additionally produce a data-driven model of the soft fish biorobot, which tracks experimental data to 1% mean error in displacement. We use this model to offer broader insight into the efficacy of eGaIn strain sensing to record biological movement of body caudal appendages in animals.}