Feeding systems use similar feed evaluation approaches for all dairy ruminants, whereas feed intake prediction and requirement assessment are species-specific. This review compares the most recent models currently used to predict dry matter intake (DMI) and energy requirements in lactating dairy small ruminants (DSR), i.e., the Small Ruminant Nutrition System (SRNS, 2004 and 2010, updated in the Ruminant Nutrition System, 2016), the NRC(2007), and the new INRA-Systali (2017) systems, to highlight differences in approach, inputs required and flexibility used. Because DSR are normally group fed, accurate prediction of feed intake is necessary to adjust diet composition according to their requirements. However, when predicting DMI for ewes and goats, the approaches used, variables considered and predicted values vary considerably. All the models used are empirical and species-specific, and use body weight and milk yield as main predictors. Variables associated with diet composition and/or type of feeds are also considered, but vary among models, while dietary particle size is not considered. The SRNS and NRC do not have specific models to predict intake on pasture, while INRA uses the Fill Unit system (i.e., based on a forage of reference) to predict DMI considering both housing and grazing conditions. For energy requirements, the SNRS uses the same model for both species, while NRC and INRA use species-specific models. The INRA (2017) largely updated these predictions for all productive functions. For maintenance energy requirements, besides some variability in the basal metabolic rate and energy units used (i.e., ME or NEL), most differences regard the additional factors affecting these requirements (e.g., age, gender, body condition, growth, previous nutrition, movement, cold and heat stress, dietary nitrogen unbalance). Differently, energy requirements for milk production and pregnancy are estimated using similar approaches and values. Overall, the feeding systems for dairy sheep and goats have markedly evolved in the last decade, by improving their ability to account for the great diversity of production systems in which they are raised worldwide