In table tennis the follow-through action after a shot is an important part of skill execution. In this experiment, we aimed to extend literature around the contextual interference effect by investigating whether the way the follow-through is organised in practice affects learning of the backhand shot in table tennis. Thirty unskilled participants were allocated to blocked-variable practice, random-variable practice or a control-constant group and aimed backhand shots towards a target following ball projection from a machine. Each group completed these shots in a pre-test, a training phase with follow-through manipulations, a post-test, and a retention test. The random-variable group improved their shot accuracy from pre-test to post-test and from pre-test to retention test (both P < 0.01, d = 1.03), whereas neither the blocked-variable nor the control-constant group displayed any change in shot accuracy. Practising the follow-through in a random-variable fashion enhanced learning of the preceding shot compared with blocked-variable practice or no follow-through instructions. The benefits of learning motor skills under conditions of high contextual interference also apply to how follow-through actions are organised. The findings are of value to coaches and suggest that instructions related to the follow-through action should be considered as well as the primary skill itself.