Climate change-induced heat stress combines two challenges: high day- and nighttime temperatures, and physiological water deficit due to demand-side drought caused by increase in vapor-pressure deficit. It is one of the major factors in low productivity of maize in rainfed stress-prone environments in South Asia, affecting a large population of smallholder farmers who depend on maize for their sustenance and livelihoods. The International Maize and Wheat Improvement Center (CIMMYT) maize program in Asia, in partnership with public-sector maize research institutes and private-sector seed companies in South Asian countries, is implementing an intensive initiative for developing and deploying heat-tolerant maize that combines high yield potential with resilience to heat and drought stresses. With the integration of novel breeding tools and methods, including genomics-assisted breeding, doubled haploidy, field-based precision phenotyping, and trait-based selection, new maize germplasm with increased tolerance to heat stress is being developed for the South Asian tropics. Over a decade of concerted effort has resulted in the successful development and release of 20 high-yielding heat-tolerant maize hybrids in CIMMYT genetic backgrounds. Via public–private partnerships, eight hybrids are presently being deployed on over 50,000 ha in South Asian countries, including Bangladesh, Bhutan, India, Nepal, and Pakistan.

Erratic rainfall often results in intermittent drought and/or waterlogging and limits maize (Zea mays L.) productivity in many parts of the Asian tropics. Developing climate-resilient maize germplasm possessing tolerance to these key abiotic stresses without a yield penalty under optimal growing conditions is a challenge for breeders working in stress-vulnerable agro-ecologies in the region. Breeding stress-resilient maize for rainfed stress-prone ecologies is identified as one of the priority areas for CIMMYT-Asia maize program. We applied rapid cycle genomic selection (RCGS) on two multiparent yellow synthetic populations (MYS-1 and MYS-2) to improve grain yield simultaneously under drought and waterlogging conditions using genomic-estimated breeding values (GEBVs). Also, the populations were simultaneously advanced using recurrent phenotypic selection (PS) by exposing them to managed drought and waterlogging and intermating tolerant plants from the two selection environments. Selection cycles per se (C₁, C₂, and C₃) of the two populations developed using RCGS and PS approach and their test-cross progenies were evaluated separately in multilocation trials under managed drought, waterlogging, and optimal moisture conditions. Significant genetic gains were observed with both GS and PS, except with PS in MYS-2 under drought and with GS in MYS-1 under waterlogging. Realized genetic gains from GS were relatively higher under drought conditions (110 and 135 kg ha⁻¹ year⁻¹) compared to waterlogging (38 and 113 kg ha⁻¹ year⁻¹) in both MYS-1 and MYS-2, respectively. However, under waterlogging stress PS showed at par or better than GS as gain per year with PS was 80 and 90 kg ha⁻¹, whereas with GS it was 90 and 43 kg ha⁻¹ for MYS-1 and MYS-2, respectively. Our findings suggested that careful constitution of a multiparent population by involving trait donors for targeted stresses, along with elite high-yielding parents from diverse genetic background, and its improvement using RCGS is an effective breeding approach to build multiple stress tolerance without compromising yield when tested under optimal conditions