From our experience observing novice animators, it is clear that setting keyframe spatial values is straightforward while specifying the keyframe timing is difficult and often time consuming. This paper presents a modified approach to the keyframing paradigm, performance timing, that lets the user focus on the timing of keyframes separately from the spatial values. In performance timing, the user “acts-out” the timing information using a simple 2D input device such as a mouse or pen-tablet. The user’s input is analyzed and features of the user’s input are mapped to the spatial features of the keyframed motion. The keyframes are then distributed in time according to the timing of the user’s input path. We demonstrate the approach on several scenes and discuss the situations in which it is most and least effective. We present the results of a user study of over 20 subjects in which we compare accuracy using performance timing to accuracy using a standard animation package for specifying keyframe timing.