Modern agriculture depends on chemically synthesized nitrogen fertilizer, which ensures high yields but also can carry significant environmental and economic costs. Biological nitrogen fixation (BNF) already supplies nitrogen to legume crops and several avenues of research are underway to extend it to non-legume crops. In maize (Zea mays), aerial roots have been shown to contribute to BNF in some varieties, and both having many aerial roots and large aerial roots contributes to the fixation trait. However, much of the genetics controlling aerial root number and size is still unknown. Here we validate and quantify BNF in maize varieties from Southern Mexico under controlled conditions and evaluate a population of double haploids derived from the elite inbred PHZ51 crossed with these varieties. We find that most aerial root traits (root number, nodes with roots, root size) are reasonably heritable (h2 0.5-0.75) and generally uncorrelated with each other. QTL mapping identifies 5 QTL each affecting nodes with aerial roots and aerial root number per node; in both cases all but 1 QTL show an increase from the landrace allele. We also identify 11 QTL for aerial root diameter, with most positive QTL coming from PHZ51. Between the two populations, only a few QTL overlap, indicating a presumably high diversity of genes affecting aerial root morphology in landrace populations. Combining the best QTL into elite material may provide a path toward meaningful levels of BNF for maize, and additional work is needed to determine how viable this approach will be in field settings.