Habitat destruction and fragmentation are the most frequently cited reasons for bringing species to the brink of extinction, raising the crucial question of how much space animal populations need to survive. The first approach to this question is to study species spectra and population development on islands of different sizes. Forest islands and patches of particular locations on the mainland are comparable to natural islands in the ocean. They can be analyzed in terms of size and distance to similar patches and correlated to the number of plant and animal taxa. Experiments with the artificial reduction of habitat islands and its consequences are also discussed. Studies on individual plant species and their populations as habitat islands are summarized for both woody and herbaceous plants. In experiments, the isolation factor can be manipulated, as isolation is crucial in a network of biotopes. Corridors connect habitats in a fragmented landscape. Studies on small-scale and transnational landscape corridors are presented. The colonization of habitat patches can also be studied experimentally, e.g., in artificial water containers.