We aimed to examine and compare sex-differences in people receiving treatment for psychotic illnesses in community settings, based on long or short duration of illness; expecting association between longer illness-duration and worse outcomes in women and men. Clinical, demographic and service-use data from the Survey of High Impact Psychosis were analysed by sex and duration of illness (≤5 years; ≥6 years), using independent t-tests, chi-square tests, one-way ANOVA, and Cramer's V. Of the 1825 participants, 47% had schizophrenia, 17.5% bipolar and 16.1% schizo-affective disorders. More women than men had undertaken post-school education, maintained relationships, and been living in their own homes. Women with a shorter-illness-duration showed social functioning equivalent to non-ill women in the general population. Men tended to have an early illness onset, show premorbid dysfunction, be single, show severe disability, and to use illicit substances. Men with a longer-illness-duration were very socially disadvantaged and isolated, often experiencing homelessness and substance use. Men with a short-illness-duration were most likely to be in paid employment, but two-thirds earned less than $AUD500 per fortnight. Men with longer-illness-duration showed most disability, socially and globally. Interventions should be guided by diagnosis, but also by a person's sex and duration of illness.