|dc.description.abstract||Increasing the concentrations of Fe and Zn in staple food crops through breeding has been proposed as one strategy to minimize the adverse effects of widespread mineral deficiencies in humans. This approach requires the presence of adequate genetic differences in concentrations of grain minerals for improvement. Eight trials involving different sets of tropical maize inbred lines adapted to the lowlands and mid-altitudes were, therefore, evaluated for concentrations of grain Fe, Zn and other minerals in two locations. The combined analyses of variance showed significant variation in concentrations of grain minerals among inbred lines in each trial, which was always greater than the variation caused by locations and line × location interactions. The line × location interaction had no significant effect on concentrations of Fe, Zn, Cu, Mg and P in at least three trials of lowland inbred lines. The line × location interaction also did not significantly affect the concentrations of any minerals, except S, in at least three trials of mid-altitude inbred lines. The best-inbred lines identified from each trial had 32–78% more Fe and 14–180% more Zn than their trial average. The first two principal component axes, which accounted for 55–64% of the total variation in kernel mineral concentrations, stratified the inbred lines in each trial into four groups based on differences in their grain mineral compositions. None of the correlations of Fe and Zn with Mn, Cu, Ca, Mg, K, P and S was significant and negative in the various trials, while the correlations of Fe with Zn were positive and significant (r = 0.55 to r = 0.68, p < 0.0001) in almost all the trials. These results suggest that a genetic potential exists for concurrent improvement of Fe and Zn without lowering the concentrations of other grain minerals in maize.