Iron deficiency and iron deficiency anemia have been shown to have negative effects on behavioral and electroencephalographic (EEG) measures of cognitive function, with many of these being reversible with timely repletion. We report here results from a randomized, controlled, double‐blind efficacy feeding trial testing the hypothesis that iron repletion would produce improvements in measures of brain function during cognitive tasks. Participants (healthy females, 18–27 yrs, n = 239) screened for low iron status, were randomized to groups that consumed either a high‐iron‐biofortified (BB) or a normal‐iron bean (CN) for 20 weeks. A subsample (n = 47) performed tests of attention and memory with concurrent EEG at baseline (BL) and endline (EL). A total of 29 participants had depleted iron stores (serum ferritin < 20 mg/L) and were non‐anemic, and 18 were iron depleted and anemic (serum ferritin < 20 mg/L and hemoglobin < 12 g/dL). We focus here on the BL and EL EEG data from the recognition memory task. Participants studied a set of images and were tested with equal numbers of studied (old) and unstudied (new) images, presented at test with 2, 3, or 4 cues. Dependent measures were the amplitude and latency of the P2 component (the earliest EEG feature that distinguishes old and new items), along with power in the θ‐ (4–7 Hz) and γ‐bands (> 30 Hz), which have been shown to distinguish old and new items and to index effortful attention in recall. Data were analyzed using 2 (time: BL, EL) by 2 (condition: CN, BB) repeated‐measures analyses of variance, run separately for new and old items. In addition, each dependent measure was regressed onto number of cues and condition, separately for new and old items at BL and EL. Significant (all p < .05) time by condition interactions were obtained for P2 latencies for the new items, and for the P2 amplitudes and latencies, θ‐band power in the central electrodes, and γ‐band power in the frontal electrodes for the old items. Increases were observed for P2 amplitudes and θ‐ and γ‐band power, decreases were observed for P2 latencies, and all changes were largest for those consuming BB. Significant (all R2 > 0.15) effects of number of cues were obtained for old items at BL for γ‐band power in frontal and central electrodes, and for old items at EL for all of the dependent measures, with the effects being largest for those consuming BB. Increasing number of cues produced increases in P2 amplitudes, decreases in P2 latencies, and increases in θ‐ and γ‐band power, with those increases being largest for those consuming BB. Taken together, these results suggest that the consumption of BB improves brain function in the service of memory.Support or Funding InformationFunded by Harvest Plus.