Abstract: Watermelon polysaccharide (SWP) was extracted from watermelon and watermelon polysaccharide sulfate (SSWP) was prepared by chlorosulfonic acid-pyridine. The sulfur content in sulfated watermelon polysaccharide as an indicator, based on response surface design (Response surface methodology, RSM) , the combination of three factors:the effect of sulfur in sulfate reaction time, temperature and esterification reagent ratio (chlorosulfonic acid:pyridine) was optimized. The reaction time of 3.2h, reaction temperature of 61℃, and the ratio of esterification of reagent (chlorosulfonic acid ∶pyridyl) which was 1∶1.3 were the optimum combination of conditions. Under this condition, the content of sulfur in sulfated watermelon polysaccharides was up to11.98%. Watermelon polysaccharide were scanned by IR spectra showing new absorption peak appeared at1250.76cm-1and 810.74cm-1, which indicated that the sulfation was successful. Its monosaccharide composition was analysed by GC-MS, which showed that watermelon polysaccharide was composed of six kinds of monosaccharide :rhamnose, arabinose, xylose, mannose, glucose, galactose, and the molar ratio was 4.29∶11.72∶8.66∶5.35∶13.62∶56.36. Its molecular weight was measured by size exclusion chromatography (SEC) , Mn=2.717×105, Mw=1.232×105. By measuring the scavenging rates of sulfated watermelon polysaccharide on DPPH radicals, hydroxyl radical and reducing power, it showed that the sulfated watermelon polysaccharides had significantly improved on scavenging rates and reducing power of hydroxyl radical and DPPH radical than the unesterified ones before.