We aimed to compare the dietary intakes of Australian patients with non-alcoholic fatty liver disease (NAFLD) to general Australian population intake data and determine whether the intake of any nutrient or food group was able to predict the degree of steatosis. Dietary data from fifty adult patients with NAFLD were compared to intake data from the Australian Health Survey for energy, macronutrients, fat sub-types, alcohol, iron, folate, sugar, fibre, sodium and caffeine. Linear regression models adjusting for potential confounders (age, sex, physical activity and body mass index) were used to examine predictive relationships between hepatic steatosis (quantified via magnetic resonance spectroscopy) and dietary components. The mean percentage differences between NAFLD and Australian usual intakes were significant for energy, protein, total fat, saturated fat, monounsaturated and polyunsaturated fats (all p < 0.001). The contribution of fat and protein to total energy intake was significantly higher in the NAFLD cohort (p < 0.05). No individual nutrients or food groups were strongly related to hepatic fat in the adjusted models. Higher overall consumption appears to be a major feature of dietary intake in NAFLD when compared to the general population. A whole-diet approach to NAFLD treatment and prevention is likely to be more effective than focusing on single food components.