To test whether reduced away-from-home food expenditure (AFHFE) and better nutrition knowledge and beliefs (NKB) are associated with dietary quality among US adults.

The dietary intake data (average of two 24 h recalls) used were collected from US adults (20–65 years) participating in two cross-sectional surveys, the 1994–96 Continuing Survey of Food Intake by Individuals (CSFII; n 7148) and the CSFII/Diet and Health Knowledge Survey (DHKS; n 4252).

Dietary quality was assessed using selected nutrients and food groups and the 2005 revised US Department of Agriculture Healthy Eating Index (HEI).

(i) Absolute AFHFE (weekly, per capita) and proportion of this exposure out of total food expenditure (relative expenditure); (ii) NKB score using a composite of an eleven-item scale elicited among the CSFII/DHKS subgroup.

We used t tests, χ2 tests, Wilcoxon rank-sum tests and multivariate linear regression models adjusting standard errors for sample design complexity. We utilized a change-in-estimate approach to assess mediation. For effect modification, we tested the significance of interaction terms (NKB × AFHFE).

Absolute AFHFE was positively associated with grams of fat (β = 0·14 (se 0·06)) and saturated fat (β = 0·02 (se 0·01)) and negatively associated with fibre (β = −0·02 (se 0·01)) and HEI (β = −0·08 (se 0·01)). Relative AFHFE mediated NKB effects on intakes such as fat, saturated fat, cholesterol, Na, and fruits and vegetables (change in estimate >10 %). Among subjects with a poor NKB score, higher AFHFE resulted in lower diet quality, particularly Na and cholesterol intakes.

Higher AFHFE was associated with a lower dietary quality and interacted antagonistically with NKB in some instances, while mediating the relationship between NKB and dietary quality in others.