The doubled haploid (DH) technology became the preferred method of inbred line production in maize and other crop species. Maize DHs are produced in vivo through crosses with maternal haploid inducers. Even though large, well-established maize breeding programs have been exploiting the technique for a while, a slower rate of adoption was observed in smaller breeding companies. The high price and/or lack of environmental adaptation of inducers available for licensing, or the poor performance of those free of cost, might explain why not all breeding programs were able to take advantage of this technique. The lack of adapted inducers is especially felt in tropical countries, where the poor performance of temperate inducers prevents their use for DH line production. Defining optimal breeding strategies for inducer development will, therefore, benefit many maize breeding programs. In chapter one, we review traits important to haploid inducers, explain their genetic basis and discuss breeding strategies for inducer improvement. In chapter two, we describe a genome-wide association study designed to identify QTL affecting the haploid induction rate (HIR) of maize maternal haploid inducers. In chapter three, we evaluate the usefulness of genomic prediction (GP) for the improvement of HIR and agronomic traits important to inducers. In this chapter, we also evaluate the utility of GP for parental selection and cross prediction. The outcomes of these researches can be directly applied to inducer improvement, whose performance has a great impact on the cost of DH line production.