Plasma ferritin (F) concentrations were measured (range 12-245 ng. ml-1) as an indicator of iron status in 24 male and 45 female healthy, nonancmic athletes. Usual food intake was assessed using a self-completed but supervised food frequency questionnaire. Linear regression was applied to determine the relationship between log F (as data was skewed) and average daily intake of iron (Fe). meat (M), protein (Pr). carbohydrate (CHO), fat (Fa), fiber (Fi), and kilojoules (Kj). There were negative correlations between F and each of Kj (P < 0.01) and CHO (P < 0.001), and a positive correlation between F and Pr/Kj (P < 0.01). No significant relationship existed between F and any of Fe, M, Fa. or Fi. These data suggest that iron or meat intake may not be important determinants of iron status in these athletes. Instead, the percentage of protein in the diet may be more influential on F, as may the (negative) effect of Kj and CHO intake, or at least their reflection of energy expenditure.