Seabirds spend most of their life at sea and have to possess a waterproof plumage to be able to sit on water for extended periods. We tracked juvenile Barau’s petrels for the first time, when they leave their birth colony and found that half of the transmitters stopped soon after they first landed on the water off Réunion Island. We suspected from observation at sea that birds may have problems with the waterproofness of their plumage. Therefore during the next season we set up a simple protocol to assess waterproofness of the plumage of the birds just before they fledge. This protocol is based on the calculation of a wettability index expressed as the mass of water logged in the plumage after simulating the bird sitting on the sea surface. We found that at least one third of chicks ready to fledge gained more than 4 g of water in 20 s, indicating that plumage was not waterproof. Within a sample of birds having fledged and before reaching the sea surface, a similar proportion of birds had their plumage not waterproof. For the birds tracked, only those with an index indicating a waterproof plumage successfully dispersed after their first touch on the sea surface. We provide evidence of a possible major cause of mortality of juvenile seabirds that has not been described previously in wild seabirds, but could exist in other species. This issue may be a major cause of threat to the endangered Barau’s petrel.