Hunger and food insecurity have been identified as core indicators of an individual's nutritional state that should be assessed in nutrition surveillance activities. Such an assessment requires a valid measure of these phenomena. This paper describes further work on the construction of measures of hunger and food insecurity based on the Radimer/Cornell items and provides an assessment of their validity. A random sample survey of 193 households with women and children living at home was conducted in 1993 in a rural county in New York State. A questionnaire containing the Radimer/Cornell hunger and food insecurity items, information about demographic characteristics, frequency of fruit and vegetable consumption, and household food supplies was administered to subjects. Measures were constructed that identified households experiencing household- and individual-level food insecurity and households with hungry children. The construct and criterion-related validity of the measures was assessed. In relation to criterion-related validity, as food insecurity worsened, there was a significant and progressive increase in the percentage of subjects participating in food programs and having low income, education and employment and a significant decline in average household food availability and fruit and vegetable consumption. These results support the validity of the Radimer/Cornell measures and illustrate their ability to differentiate among groups of households experiencing increasingly severe food insecurity and hunger.