Over semi-arid agricultural regions such as the U.S. Southern High Plains (SHP) producers of dryland crops need to know which management practices increase yields and decrease production risk. Here, a modeling approach is used to explore management options (MO) that increase dryland cotton yields and estimate those practice's yield risk effects under current SHP climate conditions. To simulate current dryland yield variability, dense distributions of lint yield outcomes were generated using the CROPGRO-Cotton crop model driven by weather inputs from 21 SHP weather stations during 2005-2016. Management effects were explored by repeating simulations over 32 MOs defined by 4 planting dates, 4 planting densities, and applying or not applying nitrogen. Both earlier planting date and decreased plant density increased median simulated yields, with earlier planting having the greatest positive yield effects. The simulated MO that produced the highest median lint yields planted on the earliest planting date (May 15), at the lowest density (3 plants m −1), and applied no nitrogen. Recent SHP field studies generally confirm the earlier planting date effect, but suggest insignificant yield effects for different seeding rates. Even so, negligible yield effects and lower input costs favor lower seeding densities from a profit standpoint. These crop simulations demonstrate a modeling-based method for climate-related agricultural risk management, and suggest mid-May planting dates and low plant densities as part of management practices that increase yields and profits in dryland SHP cotton production.