Most countries in the world depend primarily on rainfed agriculture for their grain food and there are strong reasons to believe that investments in lowyielding rainfed agriculture could have large impacts on poverty reduction. Most works in crop physiology related to breeding does focus on either yield potential or yield under stressful conditions which may reverse performance in other environments. These finding may explain why past breeding programs have largely not as expected to produce an impact on subsistence agriculture in developing countries. Correlation between the same traits in two environments may be negative or positive, depending on the environment where the experiment was grown. Crop selection in natural rainfall conditions vary in different years with additional stress-managed experiments, particularly when error variance is high and heritability estimate is low, resulted in optimum cultivar selection. These cultivar, yield better than any other available cultivar in high to low rainfall conditions, moreover an economic production under severe drought stress and therefore, increased productivity in a wide range of unpredictable rainfed environments. A researcher can use improved statistical design and analysis techniques, in multienvironments information, and consider secondary traits for making selection decisions. These alternative traits should still be much simpler than the complex genes controlling ultimately yield itself under a wide range of conditions. Earliness, canopy temperature, maintaining high kernel weight and leaf senescence are considered inherent heat and drought tolerance in wheat
