The use of semi-isolated habitats such as oceanic islands, lakes and mountain summits as model systems has played a crucial role in the development of evolutionary and ecological theory. Soon after the discovery of life in caves, different pioneering authors similarly recognized the great potential of these peculiar habitats as biological model systems. In their 1969 paper in Science, 'The cave environment', Poulson and White discussed how caves can be used as natural laboratories in which to study the underlying principles governing the dynamics of more complex environments. Together with other seminal syntheses published at the time, this work contributed to establishing the conceptual foundation for expanding the scope and relevance of cave-based studies. Fifty years after, the aim of this review is to show why and how caves and other subterranean habitats can be used as eco-evolutionary laboratories. Recent advances and directions in different areas are provided, encompassing community ecology, trophic-webs and ecological networks, conservation biology, macroecology and climate change biology. Special emphasis is given to discuss how caves are only part of the extended network of fissures and cracks that permeate most substrates and, thus, their ecological role as habitat islands is critically discussed. Numerous studies have quantified the relative contribution of abiotic, biotic and historical factors in driving species distributions and community turnovers in space and time, from local to regional scales. Conversely, knowledge of macroecological patterns of subterranean organisms at a global scale remains largely elusive, due to major geographical and taxonomical biases. Also, knowledge regarding subterranean trophic webs and the effect of anthropogenic climate change on deep subterranean ecosystems is still limited. In these research fields, the extensive use of novel molecular and statistical tools may hold promise for quickly producing relevant information not accessible hitherto.