Human behaviour consists in large parts of action sequences that are often repeated in mostly the same way. Through such extensive repetitions, these response patterns can become automatic or habitual, but our environment often confronts us with events to which we have to react flexibly and in a goal-directed manner. In order to understand how implicitly learned and automatized motor sequences interact with an interfering goal-directed task, we developed a novel behavioural paradigm in which we combined implicit sequence learning through repetition with an explicitly instructed goal-directed task. This goal-directed task requires the selection of the option with the higher reward probability, in a fast succession of trials with short response deadlines. Within a motor sequence task, participants were offered two response options and were instructed to choose the one with the higher reward probability. As expected, we found that participants learned the motor sequence through repetition and acted, in trials with two response options, in a goal-directed manner. In addition, there was evidence of a strengthening of the learned action sequence from the first to the second day of the experiment. We found a rich and telling pattern of interactions between automatic and goal-directed behavior, expressed in reaction times, choices and error rates. This novel paradigm may help to shed light on the development of automatic behavioural patterns and habits that arise through extensive repetition, and their interaction with goal-directed behaviours under time pressure.