The optimal diet composition to prevent obesity and its complications is unknown. Study aims were to determine the association of diet composition with energy intake, homeostatic model assessment-insulin resistance (HOMA-IR), and C-reactive protein (CRP). Data were from the NHANES for eligible adults aged 20-74 y from 2005 to 2006 (n = 3073). Energy intake and diet composition were obtained by dietary recall. HOMA-IR was calculated from fasting insulin and glucose concentrations, and CRP was measured directly. Changes for a 1-point increase in percentage of sugar, saturated fatty acids (SFAs), monounsaturated fatty acids (MUFAs), polyunsaturated fatty acids (PUFAs), and alcohol were determined across their means in exchange for a 1-point decrease in percentage of nonsugar carbohydrates. Regression analyses were performed, and means ± SEs were estimated. Increasing the percentage of sugar was associated with increased energy intake in men (23 ± 5 kcal; P < 0.001) and women (12 ± 3 kcal; P = 0.002). In men, increasing percentages of SFAs (58 ± 13 kcal; P = 0.001) and PUFAs (66 ± 19 kcal; P < 0.001) were associated with increased energy intake. In women, increasing percentages of SFAs (27 ± 10 kcal; P = 0.02), PUFAs (43 ± 6 kcal; P < 0.001), and MUFAs (36 ± 13 kcal; P = 0.01) were associated with increased energy intake. Increasing the percentage of alcohol was associated with increased energy intake in men (38 ± 7 kcal; P < 0.001) and women (25 ± 8 kcal; P = 0.001). Obesity was associated with increased HOMA-IR and CRP in both genders (all P ≤ 0.001). Increasing PUFAs was associated with decreasing CRP in men (P = 0.02). In conclusion, increasing the percentage of calories from sugar, fats, and alcohol was associated with substantially increased energy intake but had minimal association with HOMA-IR and CRP.