Prospective memory (PM) refers to the ability to remember to complete a task at the appropriate moment in the future. Past research has found reminders can improve PM performance in both laboratory and naturalistic settings, but few projects have examined the circumstances when reminders are most beneficial. Three experiments in the present study tested the effect of reminders in an event-based PM task under different cognitive loads. In Experiments 1 (specific targets) and 2 (nonspecific targets), load was manipulated by having participants respond to a single (low load) or multiple (high load) targets. In Experiment 3, the association between pairs was manipulated by presenting strongly associated pairs (low load) or weakly associated pairs (high load). Half of the participants in each experiment received reminders whereby PM target information was placed at the top of the computer screen. Across all three experiments, it was found that the benefit of reminders was greater under high load than low load conditions. Importantly, the improvements in PM from reminders generally occurred without any additional cost to ongoing task performance and without any reduction of retrospective memory for the targets at the end of the experiment. Together these results suggest that reminders can be beneficial for reducing PM failures, particularly under high load, without the potential downside of increased effort expenditure.