This overview describes a series of studies on the feeding behaviour of goats in the low deciduous forest (LDF) and some of the factors that affect such behaviour. Those multidisciplinary studies included field trials as well as controlled pen trials. Field observations were performed with the direct observation method (DOM) and by establishing a bite code grid adapted to the LDF, and the pen studies used a cafeteria strategy using leaves harvested from different plants of the LDF. Studies exploring the selection of feed resources used different selection indexes. Field observations showed that goats consumed >90 plant species from the heterogeneous vegetation, but only a few (7 - 15 species) represented >90% of the daily intake, and that differed between rainy and dry seasons. Although the LDF vegetation allows goats to match their crude protein requirements in both seasons, dietary energy was the limiting factor, and the condensed tannin (CT) content of the harvested diet was always high. Field and cafeteria studies suggested that goats optimize their diet with the resources available as they prefer to consume leaves with lower fibrous content while plant secondary metabolites (PSM) such as condensed tannins (CT) seemed less influential. Field and pen studies confirmed that CT blocking with polyethylene glycol did not modify the intake and selection of the CT-rich foliage which could be related to tannin-binding salivary proteins which were identified as an innate feature of the local goats. While pen studies showed that browsing experience helped goats select feeds of better digestibility and lower PSM content, helping them optimize their nutrition, field studies showed that adult goats consumed more plant species, performed larger bites, at higher strata than kids, while the proportion of life forms consumed was similar. Most field and pen studies showed that goats generally do not consume CT-rich foliage as a therapeutic self-medication strategy against gastrointestinal nematodes (GIN), instead, a prophylactic strategy seemed feasible. This behavioural feature could aim at different goals: (i) to avoid excess nitrogen in the rumen, (ii) to maintain the CP:energy balance (reducing energy loss due to urea urinary excretion), and (iii) to avoid exsheathment of GIN L3 (affecting their life-cycle). The DOM showed that goats consume small quantities of plants containing toxic compounds and they shift from browsing early in the morning to grazing some hours later, and both strategies warrant further research. The DOM also helped to relate plant life forms consumed with the degree of GIN infection, which deepens our understanding of GIN epidemiology in heterogeneous vegetation ecosystems. From these findings on the feeding behaviour of goats in the LDF, new research opportunities were proposed.