Recent behavioural research has investigated whether viewing someone perform an action results in activation of that action by the observer. Postulated empirical support for this 'ideo-motor (IM) conjecture' typically rests upon two types of experimental paradigm (reaction time and movement tracking tasks). These paradigms purport to show movement facilitation when compatible movements are observed and vice versa, but only for biological stimuli. Unfortunately, these paradigms often contain confounding (and unavoidable) generic stimulus-response compatibility effects that are not restricted to observed human movement. The current study demonstrates in three experiments that equivalent compatibility effects can be produced by non-biological stimuli. These results suggest that existing empirical paradigms may not, and perhaps cannot, support the IM-conjecture.