To investigate the prevalence of food insufficiency and factors associated with it, two questions assessing household and individual food insufficiency were included in 13 regional health surveys conducted in Queensland in 1993. The surveys used computer-assisted telephone interviewing methodology. Of the 10,451 people interviewed, 9.7 per cent and 6.4 per cent reported household and individual food insufficiency, respectively, and 11.3 per cent reported at least one type. Prevalence was significantly higher in women than men and in urban than rural residents, and decreased monotonically with increasing age from 16.6 per cent in 18- to 30-year-olds to 1.7 per cent in over-70-year-olds. Higher prevalence also was associated with lower income, unemployment, single or separated, divorced or widowed status versus married (or de facto), one-adult households, and shared accommodation. Lower prevalence was associated with more education in those aged 50 and under but not in those over 50 years. Using logistic regression to control simultaneously for important sociodemographic factors, we found that risk of food insufficiency was most highly associated with age and income (threefold risk), unemployment and shared accommodation (twofold risk) and one-adult households, and being single versus separated, widowed or divorced (one-and-a-half-fold risk). Some differences in risks existed between men and women and between rural and urban residents, although none excluded the role of chance. Association of the items with lower reported fruit, vegetable and meat intake, poorer health status, and greater underweight supports their validity.