Decisions about actions are thought to arise from a biased competition between simultaneously activated movement plans. In support, reaction times (RT) are longer in situations of increased competition, such as during incongruent trials in Flanker tasks or equiprobable target locations (point of subjective equality; PSE) in hand selection tasks. Still, uncertainty has mostly been studied in contexts in which the target goal is instructed or in which overt execution of both responses is possible, limiting the evidence for a lower-level, action-based competitive process evolving within a single cortical hemisphere. Therefore, we aimed to investigate the effect of decisional uncertainty on RTs in a free choice task where overt execution of both responses was impossible. We developed an obstacle avoidance task in which participants (n = 11), using their dominant hand, had to reach toward a visual target while avoiding an obstacle positioned along the movement path. The target was positioned straight ahead of participants and remained unchanged across trials. Uncertainty was manipulated by placing a 9cm obstacle on participants’ direct path to the target (at the PSE) (maximal uncertainty condition), or an 18cm obstacle shifted laterally by ±4.5cm from the PSE location (minimal uncertainty condition). Importantly, this setup equated movement trajectories across conditions, allowing us to isolate the influence of uncertainty on RTs. Results revealed RTs were significantly slower in the maximal uncertainty condition (375ms ± 32) as compared to the minimal uncertainty condition (355ms ± 28; p=0.0005). This suggests that a competitive process also underlies action-based free-choice decisions.