CIMMYT’s Global Wheat Program has improved drought adaptation using strategic crosses that combine complementary physiological traits (PT) as demonstrated by a new generation of lines distributed in the 17^th SAWYT (Semi-Arid Wheat Yield Trial); average performance of PT lines was superior to conventional lines at 75% of international sites in 2010. To unravel the genetic basis of PTs, populations were designed to control phenology, including Seri x Babax recombinant inbred lines (RILs) and Wheat Association Mapping Initiative (WAMI) panel. However, markers linked to QTL are not routinely used for selection and pyramiding of PTs. We will discuss factors that limit adoption of MAS as well as the tradeoffs between molecular marker and phenotyping approaches. For example, CIMMYT uses two breeding cycles per year, the winter cycle (Cuidad Obregon) permits expression of yield potential and adaptation to a range of drought and heat stresses. Phenotyping in this environment is likely to remain the most effective selection strategy due to . On the other hand, use of markers has greater potential in the summer cycle (Toluca and El Batan) where selection for yield and stress adaptive traits is not feasible. Recent changes in low cost and high-throughput genotyping will boost MAS. The availability of gene-based marker panels are an attractive prospect, in particular for early selection for flowering time alleles (Ppd and Vrn), and plant height (Rht). Association mapping efforts (e.g. WAMI) identify useful genetic regions across a wide range of genetic backgrounds and are more likely to have higher generic value than markers identified in bi-parental populations. The potential of genomic selection also will be discussed.