Landscape linkages are hypothetical objects developed in geographic information systems (GIS) proposed to connect areas of habitat in fragmented landscapes. Assuming they are used by organisms in nature, linkages could be an important tool for biodiversity conservation. However, large-extent connectivity models (e.g., at the US state scale) are generally not based on evidence of wildlife occurrence and testing whether or not wild animals follow these artificial pathways created by conservation planners have given mixed results. Using >180,000 wildlife detections over 20 years, we evaluated potential utility of five California linkage models for common California mammal, reptile and amphibian species in two ways: 1) occupancy modeling and 2) roadkill detections as a proxy for wildlife movement. We found that linkage areas were not important predictors for the probability of landscape occupancy for most species. Linkage areas were also not consistent predictors of conflict on roads. These results show that hypothetical landscape linkages are not an all-purpose conservation strategy. Unless validated using data from wildlife occurrence, caution should be exercised when using linkages in land and transportation-based conservation planning. Connectivity as a gradient across the landscape should be the target of conservation, including in land-use and transportation planning.