The paradigmatic transgender woman is often negatively oversexualised, pornographised and fetishised in mainstream conceptualisations and discourses. However, self-sexualisation by transgender individuals is often portrayed as a (sex-)positive social phenomenon. Little research has been conducted that analyses the self-sexualisation strategies of the multiple instantiations of gender-variant identity, including transmasculine and non-binary social actors. This paper uses a corpus-informed socio-cognitive approach to critical discourse studies to identify differences between the self-sexualisation strategies and underpinning cognitive models of different gender-variant user-groups on Twitter. 2,565 users are coded into five categories: (1) transfeminine; (2) transmasculine; (3) transsexual; (4) transvestite; (5) non-binary. Findings show that transvestite- and transsexual-identifying users most closely fit the pornographised and fetishised conceptualisation, whilst non-binary users are the least self-sexualising user-group.