The importance of landscape connectivity in determining biodiversity outcomes under environmental change has led to indices of connectivity becoming amongst the most widely used measures in conservation. Thus, it is vital that our understanding of connectivity and our use of indices describing it are reliable. Dispersal is the key ecological process involved in determining connectivity, and there is increasing evidence of substantial within-population variability in dispersal behaviours. Here, we incorporate this inter-individual variability into two approaches for estimating connectivity, least cost path analysis and stochastic movement simulation. Illustrative results demonstrate that including dispersal variability can yield substantially different estimates of connectivity. While connectivity is typically similar between nearby patches, the frequency of movements between patches further apart is often substantially increased when inter-individual variability is included. Given the disproportionate role that unusual long-distance dispersal events play in spatial dynamics, connectivity indices should seek to incorporate variability in dispersal behaviour.