<p>We investigated how separately and sequentially acquired visual and motor experiences shape action prediction mechanisms. There is evidence that physical practice leads to an implicit, motor-based prediction process, compared to visual practice, which is more strategic. However, the relative dominance or flexibility of these mechanisms is not well understood. Here we used a motor secondary task paradigm to evaluate effector-specific interference in action predictions, which has previously given evidence for a motor simulation-based prediction process after physical practice. Participants across two groups (<i>N</i> = 40) received both isolated and sequentially combined motor and visual practice across two days; either throwing darts to three different sections of a dartboard “motor” or watching and predicting outcomes of occluded throws “visual”. The Motor-to-Visual group threw on Day 1 and watched on Day 2 and the Visual-to-Motor group did the reverse. Prediction tasks were performed pre and post practice each day, with some trials involving motor secondary tasks, performed with the observed (right) or non-observed hand. Consistent with previous work, the Motor-to-Visual group after physical practice improved prediction accuracy on Day 1, except when performing the secondary task with their right hand. After visual practice on Day 2, prediction accuracy was maintained, but without secondary task interference. The Visual-to-Motor group also improved predictions, but with no secondary task interference on either day, resulting in greater accuracy overall. These data support the suggestion that separately acquired motor and visual experiences either allow flexibility in prediction strategies or lead to a dominance of the visually-acquired strategy.</p>