"Flexibility" tests are traditionally performed voluntarily relaxed by rotating a joint slowly; however, functional activities are performed rapidly with voluntary/reflexive muscle activity. Here, we describe the reliabilities and differences in maximum ankle range of motion (ROMmax) and plantar flexor mechanical properties at several velocities and levels of voluntary force from a new test protocol on a commercially available dynamometer. Fifteen participants had their ankle joint dorsiflexed at 5, 30, and 60 degrees s(-1) in two conditions: voluntarily relaxed and while producing 40% and 60% of maximal eccentric torque. Commonly reported variables describing ROMmax and resistance to stretch were subsequently calculated from torque and angle data. Absolute (coefficient of variation (CV%) and typical error) and relative (ICC2,1) reliabilities were determined across two testing days (>= 72 h). ROMmax relative reliability was good in voluntarily relaxed tests at 30 and 60 degrees s(-1) and moderate at 5 degrees s(-1), despite CVs <= 10% for all velocities. Tests performed with voluntary muscle activity were only reliable when performed at 5 degrees s(-1), and ROMmax reliability was moderate and CV <= 8%. For most variables, the rank order of participants differed between the slow-velocity, relaxed test, and those performed at faster speeds or with voluntary activation, indicating different information. A person's flexibility status during voluntarily relaxed fast or active stretches tended to differ from their status in the traditional voluntarily relaxed, slow-velocity test. Thus, "flexibility" tests should be completed under conditions of different stretch velocity and levels of muscle force production, and clinicians and researchers should consider the slightly larger between-day variability from slow-velocity voluntarily relaxed tests.
