To get a quick and accurate hail damage simulation method for cotton, two-year experiments about two different simulation methods were conducted at the initial bloom stage in 2017 and 2018, respectively. Method one: five defoliation treatments (0%, 25%, 50%, 75%, and 100%) were conducted in 2017; method two: canopy height was decreased by 0%, 30%, 60%, and 90% by self-made tools in 2018. Field design was a randomized complete block with three replications. Canopy coverage (CC), leaf area index (LAI), leaf dry weight (LDW), leaf mass per unit area (LMA) and yield were obtained after damage. Results showed that LAI, LDW, CC, LMA and yield decreased with the increasing damage level in two simulations, yield decreased significantly when damage level was severe. Comparing two simulated hail damages with natural hail damage, it was shown that with the increasing damage level, the variations of LDW, CC, and yield (take LAI as an independent variable) for method two were closer to natural hail damage than those for method one. These results indicate that method two is more appropriate than method one for simulation of hail damage to cotton, which could be helpful for farmers and insurance companies to estimate potential economic losses.