To measure the occurrence and correlates of food insufficiency among HIV-infected crack-cocaine users in Atlanta and Miami, USA.

Non-probability cross-sectional sample.

Inner-city hospitals in Atlanta and Miami.

Two hundred and eighty-seven HIV-infected crack users.

One-third (34 %) of respondents experienced food insufficiency within 30 d of interview. Increased odds of food insufficiency was associated with current homelessness (adjusted OR = 3·78, 95 % CI 1·70, 8·41), living alone (adjusted OR = 2·85, 95 % CI 1·36, 5·98), religious service attendance (adjusted OR = 2·34, 95 % CI 1·02, 5·38) and presence of health insurance (adjusted OR = 2·41, 95 % CI 1·06, 5·54). Monthly income greater than $US 600 (adjusted OR = 0·19, 95 % CI 0·06, 0·58) was associated with decreased odds of food insufficiency, and less than weekly crack use was marginally associated with decreased odds of food insufficiency (adjusted OR = 0·39, 95 % CI 0·13, 1·08).

Food insufficiency is very prevalent among HIV-infected urban crack-cocaine users in Atlanta and Miami. Correlates of food insufficiency confirm the social vulnerability of these individuals. Routine assessment for food insecurity should become a routine component of treatment and prevention programmes in at-risk populations.