When an action contingently yields a predictable effect, we form bi-directional action-effect associations that allow us to anticipate both the location and timing of our actions’ effects. This is evident in anticipatory eye movements towards the future effect’s location which are performed earlier when the effect’s delay is short rather than long. Such anticipatory eye movements reflect a proactive process of effect monitoring which prepares a comparison of expected and actual effects. Here, in two online eye tracking experiments, we manipulated effect locations (spatially compatible vs. incompatible in one half) and effect delays (short vs. long) to determine whether in-laboratory effects could be reliably replicated online using participants’ individual webcams. Extending prior research, we further compared irrelevant (Experiment 1) to relevant effects (response to effect feature; Experiment 2). In contrast to prior in-laboratory studies, participants anticipatorily looked towards future effects above chance only when effects were relevant. Post-experiment questions suggested that online-participants intentionally ignore irrelevant information to optimize task performance. Nevertheless, replicating in-laboratory experiments, both for relevant and irrelevant effects, participants’ first saccade towards a future effect occurred earlier for the short rather than the long effect delay. Thus, we demonstrate that anticipatory eye movements reflecting a time-sensitive proactive effect monitoring process can reliably be assessed both in-laboratory as well as online. However, when investigating anticipatory saccade frequencies online, additional aspects like effect relevance have to be considered.