Peripheral cueing tasks can be used to measure reflexive (automatic) attention. In these tasks, increases in response time or RT (costs) typically follow contralateral (invalid) cues as attention must move from the location of the cue to the target. Reductions in RT (benefits) to a target typically follow ipsilateral (valid) cues because the cue draws attention to where the target will appear. Two exceptions to RT benefits are inhibition of return (IOR) and masking. IOR is the tendency to respond slower to targets that appear in locations attended within the last 200–2000 ms. Masking occurs when the visibility of a target is blocked by another stimulus (e.g., the cue). Herein, we describe two experiments, both using a modified Posner task with “earth rockets” as cues and “alien spaceships” as targets. Cues were equally likely to appear on the left or right side of a display following targets. Participants were instructed to press a left or right key corresponding to a left or right target. In Experiment 1, we obtained data from 203 children (10.58–16.55 years old). We discovered unexpected costs following cues that typically provide RT benefits. In Experiment 2, we explored IOR, masking, and age differences in the occurrence of these costs. We manipulated the cue-target temporal distance (“stimulus onset asynchrony” or SOA) to explore IOR and the cue-target spatial distance to explore masking. We also considered a wider age range. Sixty-three children and 41 young adults participated. Experiment 2 revealed a three-way interaction between SOA, spatial distance, and age. At the shorter SOA (100 ms) and moderate spatial distance, unexpected costs followed valid cues for younger children (7.07–10.15 years old). These costs also occurred in young adults (18.00–23.02 years old) following far distance cues at this SOA. At the longer SOA (200 ms), these costs followed moderate and far cues for younger children and near cues for young adults. Older children (10.31–14.92 years) did not have unexpected costs. We explain the findings in terms of masking, IOR, and possible developmental mechanisms.