Background: Ovarian cancer is ranked as the fifth most common cause of cancer death in women. In Malaysia, it is the fourth most common cancer in females. CA125 has been the tumor marker of choice in ovarian cancer but its diagnostic specificity in early stages is only 50%. Hence, there is a critical need to identify an alternative tumor marker that is capable of detecting detect ovarian cancer at an early stage. HE4 is a new tumor marker proposed for the early diagnosis of ovarian cancer and disease recurrence. Currently, none of the normal ranges of HE4 quoted in the literature are based on data for a multiethnic Asian population. Therefore, the aim of this study was to determine reference intervals for HE4 in an Asian population presenting in University Malaya Medical Centre, a tertiary reference hospital. Materials and Methods: 300 healthy women were recruited comprising 150 premenopausal and 150 postmenopausal women, aged from 20-76 years. All women were subjected to a pelvic ultrasonograph and were confirmed to be free from ovarian pathology on recruitment. Serum HE4 levels were determined by chemiluminescent microparticle immunoassay (CMIA, Abbott Architect). The reference intervals were determined following CLSI guidelines (C28-A2) using a non-parametric method. Results: The upper limits of the $95^{th}$ percentile reference interval (90%CI) for all the women collectively were 64.6 pmol/L, and 58.4 pmol/L for premenopausal) and 69.0 pmol/L for postmenopausal. The concentration of HE4 was noted to increase with age especially in women who were more than 50 years old. We also noted that our proposed reference limit was lower compared to the level given by manufacturer Abbott Architect HE4 kit insert (58.4 vs 70 pmol/L for premenopausal group and 69.0 vs 140 pmol/L in the postmenopausal group). The study also showed a significant difference in HE4 concentrations between ethnic groups (Malays and Indians). The levels of HE4 in Indians appeared higher than in Malays (p<0.05), while no significant differences were noted between the Malays and Chinese ethnic groups. Conclusions: More data are needed to establish a reference interval that will better represent the multiethnic Malaysian population. Probably a larger sampling size of equal representation of the Malay, Chinese, Indians as well as the other native ethnic communities will give us a greater confidence on whether genetics plays a role in reference interval determination.