A summary of an extensive literature survey deals with the variables known or suspected to affect the retention of learned motor behaviors over lengthy no-practice intervals. Emphasis was given to research conducted by or for the military. The variables that may affect the retention of motor skills were dichotomized into task variables and procedural variables. The task variables that may underlie the long-term retention of motor skill include (1) duration of the no-practice period, or retention interval; (2) nature of the response required to accomplish a particular motor task; (3) degree to which the learner can organize or impose order upon the elements that define the task; (4) structure of the training environment; and (5) initial or "natural" ability of the learner in performance of a task without prior practice. The procedural variables that may affect the long-term retention of motor skill include (1) degree of proficiency attained by the learner during initial training; (2) amount and kind of refresher training; (3) transfer of skills on one task to performance on another task; (4) presence of interfering activities; (5) distribution of practice during training; (6) use of part-task versus whole-task training methods; and (7) introduction of extra test trials prior to final testing. (Author/CSS)