The aim of this paper was to evaluate gender differences in the language used in United Nations (UN) General Assembly debates by one male and one female representative each from India, China, the USA, and Indonesia. The critical discourse analysis (CDA) framework of van Dijk (2015) was used along with the 25 discursive devices in this framework. The data on frequencies of variables were analysed statistically. No significant differences were observed between positive comments about oneself, negative comments about others, length of speech (in words), and the various devices by the gender of the speaker at the 0.05 level of significance. However, gender differences in terms of higher levels for males than for females were observed for populism (t(6)= 2.354, p= 0.057), norm expression (t(6)= 2.171, p= 0.073), and positive comments about oneself (t(6)= 2.224, p= 0.068). The correlation coefficients were significant between genders for positive-self, norm expression, and national self-glorification only. These results could be attributed to the small size of four male and four female speeches, leading to high values of standard error, which reflects the significance of differences. Also, there could be overlapping and mixing of gender characteristics in connection with different contexts and occasions due to emergent identities. These political identities were created in particular situations or contexts, and it was not clear how much they could be a result of female or male politicians' styles. This explanation by Sivric and Jurcic (2014) seems to be valid for this research as well. Small samples could be a limitation of this study. However, many other works used even smaller samples. The generalisability of these findings needs to be tested using other similar studies.