The present study applied a 2 × 2 experimental design to assess prospective memory (PM) development across preschool age and to examine the effect of task instructor status (researcher vs. significant other) on PM performance in 80 preschool children. Participants were required to name pictures (ongoing task [OT]), and to remember to refrain from naming but instead give a different response to certain target cues (PM task). Although the OT was of comparable difficulty for both age groups (as indicated by no performance differences), results still indicated significantly higher PM performance in 5-year-olds than in 3-year-olds, confirming the age-related increase of PM capacities between 3 and 5 years. Furthermore, results showed a performance-enhancing effect of significant others as task instructors on both age-groups. Post-hoc analysis revealed that 3-year-olds instructed by a significant other still performed marginally worse than 5-year-olds instructed by a researcher, underlining the finding that substantial changes of PM capacities take place during early childhood.