Currently it is estimated that about 1 billion people globally have non-alcoholic fatty liver disease (NAFLD), a condition in which liver fat exceeds 5 % of liver weight in the absence of significant alcohol intake. Due to the central role of the liver in metabolism, the prevalence of NAFLD is increasing in parallel with the prevalence of obesity, insulin resistance and other risk factors of metabolic diseases. However, the contribution of liver fat to the risk of type 2 diabetes mellitus and CVD, relative to other ectopic fat depots and to other risk markers, is unclear. Various studies have suggested that the accumulation of liver fat can be reduced or prevented via dietary changes. However, the amount of liver fat reduction that would be physiologically relevant, and the timeframes and dose–effect relationships for achieving this through different diet-based approaches, are unclear. Also, it is still uncertain whether the changes in liver fat per se or the associated metabolic changes are relevant. Furthermore, the methods available to measure liver fat, or even individual fatty acids, differ in sensitivity and reliability. The present report summarises key messages of presentations from different experts and related discussions from a workshop intended to capture current views and research gaps relating to the points above.