Many studies invoke moment of inertia (I[subscript ii]) as necessary and sufficient information to perceive cylinder length via wielding, yet some assert that I[subscript ii] is neither necessary (weight, "m", or static moment, M, are sufficient) nor sufficient for length judgments ("m" or M is necessary). Mathematical expressions for I[subscript ii] not involving "m" or M imply length, so I[subscript ii] could be sufficient for cylinder length judgments. In 5 experiments (N = 113), only longitudinal I[subscript ii] (the smallest principal value) informed cylinder length estimates in a novel task, rolling. Experiment 1 yielded reliable length judgments. Varying diameter supported length scaling (Experiment 2); similar length estimates occurred with both rolling and wielding (Experiment 3); feedback improved rolling length judgments to levels commensurate with wielding (Experiment 4); and length judgments shifted predictably with added mass (Experiment 5). Contrary to proposals in the literature, I[subscript ii] is sufficient information for cylinder length judgments, absent "m" or M information, even when I[subscript ii] is of minimal magnitude and the task quite novel. (Contains 2 tables, 7 figures and 3 footnotes.)