Epiphytic orchids are fundamental elements of the dynamics and composition of tropical ecosystems; there are an estimated 19,000 species worldwide, of which 85 occur in the dry forests of the basins of the Cauca and Dagua rivers in the Valle del Cauca Department in southwestern Colombia. These plants form the interface between the forest and the atmosphere and depend on aerial nutrient sources, rain, and water vapour for survival. This physiological dependence on the climate makes them especially sensitive to changes in the atmosphere and temperature, so they are ideal models for the study of climate change (CC). The objective of this study was to detect changes in the spatial distribution of seven orchid species in the tropical dry forest in the department of Valle del Cauca and their significance in terms of conservation planning for CC. A maximum entropy algorithm was used for modelling, and nine variables were analysed. Presence data for 30 municipalities came from 31 field trips, herbarium data, and the literature, and the current potential distribution was compared against the SRES-A2 scenario developed by the Intergovernmental Panel on Climate Change (IPCC) and modelled for the 2080–2100 time horizon. For the set of seven species, the results show an altitudinal increase under the future CC scenario compared to the present, but the responses vary amongst taxa, elevation, and location, depending on the degree of thermal specialization. Under the future CC scenario, the suitability of mid-mountain areas will increase at the expense of the basal areas where dry forest orchids are currently found, and the Cordillera Occidental will have a greater concentration of suitable areas than the Cordillera Central. Variables such as accessibility, land coverage, temperature, and water availability explain 88.6% of the model. A strategy to combat the impending loss of biodiversity due to CC is the establishment of Altitudinal Migration Corridors (AMCs) that connect the forest relics of the alluvial plain with the mid-mountain areas. Areas with a probability of species occurrence greater than P=0.75 have been identified with MaxEnt software, and these areas constitute “thermal refugia”, which, together with existing protected areas, form the backbone of this conservation strategy. Protection of xeric shrublands and the appropriate management of phorophytes would not only facilitate the dispersion processes of these orchids but also the survival of other flora and fauna in the dry forest of the Valle del Cauca River against CC.