Habitat loss is the greatest threat to large carnivores around the world. Maintenance of functional connectivity in fragmented landscapes is important for long-term species persistence. Here, we merge landscape genetics analyses and spatially-explicit simulations to understand future persistence and extinction of tigers (Panthera tigris) in Central India. Tigers in this landscape are restricted to Protected Areas (PAs) and forest fragments embedded within a mosaic of agricultural fields and human settlements. We examined current population connectivity of tigers across nine reserves (using 116 non-invasively sampled individuals and 12 microsatellites). Genetic data was used to infer resistance-to-movement. Our results suggest that dense human settlements and roads with high traffic are detrimental to tiger movement. We used landscape genetic simulations to model 86 different scenarios that incorporated impacts of future land-use change on inferred population connectivity and extinction. Our results confirm that genetic variability (heterozygosity) will decrease in the future and small and/or isolated PAs will have a high risk of local extinction. The average extinction risk of small PAs will reduce by 23–70% if a 5 km buffer is added around existing boundaries. Unplanned development will result in 35% lower heterozygosity and 56% higher average extinction probability for tigers within protected areas. Increasing tiger numbers in such a scenario will decrease extinction probability just by 12% (from 56% to 44%). Scenarios where habitat connectivity was enhanced and maintained, stepping-stone populations were introduced/maintained, and tiger numbers were increased, led to low overall extinction probability (between 3 and 21%). Our simulations provide a means to quantitatively evaluate the effects of different land-use change scenarios on connectivity and extinction, linking basic science to land-use change policy and planned infrastructure development.