Estimates of the incidence of sexual coercion in men's prisons are notoriously variable and fraught with conceptual and methodological problems. In 2006-2007, we conducted a computer-assisted telephone survey of a random sample of 2,018 male prisoners in New South Wales and Queensland. Of 2,626 eligible and available inmates, 76.8% consented and provided full responses. We asked about time in prison, sexual experience, attraction and (homo/bi/heterosexual) identity, attitudes, sexual contact with other inmates, reasons for having sex and practices engaged in, and about sexual coercion, including location and number of perpetrators. Most men (95.1%) identified as heterosexual. Of the total sample, 13.5% reported sexual contact with males in their lifetime: 7.8% only outside prison, 2.8% both inside and outside, and 2.7% only inside prison. Later in the interview, 144 men (7.1% of total sample) reported sexual contact with inmates in prison; the majority had few partners and no anal intercourse. Most did so for pleasure, but some for protection, i.e., to avoid assault by someone else. Before incarceration, 32.9% feared sexual assault in prison; 6.9% had been sexually threatened in prison and 2.6% had been sexually coerced ("forced or frightened into doing something sexually that [they] did not want"). Some of those coerced reported no same-sex contact. The majority of prisoners were intolerant of male-to-male sexual activity. The study achieved a high response rate and asked detailed questions to elicit reports of coercion and sex separately. Both consensual sex and sexual assault are less common than is generally believed.