Aims To compare the efficacy and safety of 3% diquafosol ophthalmic solution with those of 0.1% sodium hyaluronate ophthalmic solution in patients with dry eye in China and Singapore. Methods A total of 497 patients with dry eye (Schirmer's test, 5 mm; fluorescein and RB score, 3 points) from China and Singapore were randomised to receive either diquafosol ophthalmic solution (diquafosol) or sodium hyaluronate ophthalmic solution (HA) at 1:1 ratio. The fluorescein staining scores and rose bengal (RB) subjective symptom scores and tear film breakup time were evaluated before treatment and 2 and 4 weeks after start of treatment. Results In the diquafosol group, changes in fluorescein and RB scores compared with baseline at week 4 or at the time of discontinuation were -2.1 +/- 1.5 and -2.5 +/- 2.0, respectively. Compared with the HA group, changes in fluorescein score were non-inferior and changes in RB score were superior (p=0.019). In addition, diquafosol and HA improved tear film breakup time by 1.046 +/- 1.797 and 0.832 +/- 1.775 s, respectively (no significant intergroup difference). Adverse event onset rates were 16.3% (40 of 246 subjects) and 10.0% (25 of 251 subjects) in the diquafosol group and HA group, respectively, with borderline significant intergroup differences (p=0.046), while adverse drug reaction incidence rates were 12.2% (30 of 246 subjects) and 6.0% (15 of 251 subjects), respectively (p=0.019). Only mild adverse drug reactions (>2%) in the form of eye discharge, itching or irritation were observed. Conclusions Diquafosol improved fluorescein staining score in a manner similar to HA, and significantly improved RB score compared with HA.