Wheat is the most important cereal crop worldwide and ranks third in Nepal. Improvements in wheat yield can be done effectively by selection for yield attributing traits. In this experiment, twenty wheat genotypes were evaluated in the terai region of Nepal at Paklihawa, Rupandehi in Alpha lattice design under heat stress conditions. The characters were evaluated to find their correlation and direct and indirect effects on yield. Positive significant correlation of grain yield with No. of spikes m-2 (0.405) and harvest index (0.647) were found whereas Spike weight (-0.322) showed a significant negative correlation with grain yield. Similarly, Path analysis showed that the Harvest index (0.5511) and No. of spikelets per spike (0.3365) had a high direct effect, whereas Thousand kernel weight, Spike m-2, and Plant height showed a lower positive direct effect on grain yield. Ten spikes weight, spike length, and No. of grains per spike showed low negative direct effects. The conclusions drawn from this analysis can be useful for breeding programs under heat stress by providing information on which characteristics significantly affect the yield. However, multi-locations and multi-year trials need to be done for further verifications on the selection of such traits for improving yield.