One hundred and five college students made estimates of the lengths of freely swinging pendulums mounted in an apparatus that masked all but the top few inches of the pendulum string. The law of pendulum motion shows that visible aspects of pendulum motion uniquely specify the length of the pendulum. Thus, information is available in the display that does, if observers are able to use it, allow accurate estimation of length. In three studies estimates were found to be linear functions of actual lengths, though with wide differences in slopes among individual observers. These results, together with statements made during post-experimental interviews, are interpreted as showing that observers use a rule to the effect that length is a linear function of ‘speed’, where speed appears to be a function of both period and angular velocity.