Migratory amphibians are at high risk of negative impacts when roads intersect their upland and breeding habitats. Road mortality can reduce population abundance, survivorship, breeding, recruitment, and probability of long-term persistence. Increasingly, environmental planners recommend installation of under-road tunnels with barrier fencing to reduce mortality and direct amphibians towards the passages. Often, the permeability of these barrier and passage systems to amphibian population movements are unknown. We studied the movements of California tiger salamanders (CTS: Ambystoma californiense) in relation to solid and mesh barrier fencing attached to a 3-tunnel system between upland and breeding habitats in Stanford, California. We deployed active-trigger cameras along the fencing, used pattern recognition software to identify individuals by their unique spot patterns, and calculated individual salamander movement distances, speed, direction changes, and “success” at reaching the tunnel system. We found that migrating adult CTS moved an average of 40 m along barrier fencing before turning back into the habitat or “giving-up”. This short distance, in comparison to long migratory movements, may be explained by the orientation mechanisms salamanders use to reach their breeding sites. The probability CTS found a passage decreased rapidly with increasing distance from the tunnel system, particularly if individuals turned the “wrong” way after encountering the fence. Salamanders changed directions more often and spent more time along mesh fencing. Our results suggest that a maximum of 12.5 m between passages along CTS migration routes should allow approximately 90% of adult salamanders to encounter road crossings. Additionally, use of solid fencing or a visual barrier on mesh fencing may help to lead salamanders to passages most efficiently. These considerations can assist those seeking to design effective road mitigation for CTS and other migratory amphibians.