Five experiments examined the ability of observers to control the direction of rotation of parallel (Experiments 1-4) and polar (Experiment 5) projections of transparent objects using a control strategy based on the idea that attended surfaces are given front/convex default interpretations. Experiments 1 and 2 measured observers’ degree of control and evaluated the role of attention in the control strategy. Experiment 3 examined whether attentional constraints limit the use of the strategy with dual rotating objects. Experiment 4 measured control with unambiguous stimuli in which the direction of rotation was specified by occlusion or proximity luminance covariance, and Experiment 5 measured the control of structure (rigid, non-rigid) as well as motion by examining the control of rotating displays presented in polar perspective. The General Discussion evaluates several hypotheses concerning the nature and rationale of an attentional bias in surface interpretation.