Measuring teeth is a useful adjunct in sex assessment. Canines, in particular, have the greatest degree of sexual dimorphism, resist disease and survive postmortem trauma, rendering them highly valuable in identification. Hence, their exclusive use in odontometric sex assessment using the Mandibular Canine Index (MCI) has been advocated before. The MCI is derived as the ratio of the mesiodistal (MD) dimension of canines and the inter-canine arch width. This study has tested the use of the MCI in assessing sex on a sample from Nepal and compared its accuracy to that of absolute canine measurements. Measurements were obtained from one hundred-and-seventeen dental stone casts that belonged to 63 males and 54 females, all young adults in the age-group 19-28 years. Independent samples t-test revealed no significant sexual dimorphism in the MCI. In addition, discriminant analysis of the MCI also had poor ability to differentiate the sexes. In contrast, the absolute canine measurements revealed statistically significant male-female difference and superior ability to differentiate sex using discriminant analysis. The poor ability of the MCI in sex assessment is attributed to it being a relative value-it is obtained as the ratio of two absolute measurements (MD dimension of canines and inter-canine arch width) and does not reflect sex differences that exist in the absolute measurements per se.