It has traditionally been supposed that plantation creoles have emerged in the colonies in which they are currently spoken. However, longstanding questions suggest another approach. For one, the substrate language contributions in plantation creoles almost never correspond meaningfully to the languages spoken by the enslaved people who were imported to the colony. Second, historical evidence shows plantation creoles already existing in many colonies where Black people did not yet outnumber whites, generally considered a necessary precondition to the emergence of a creole language. Third, creoles failed to emerge in one colony after another under the Spanish, in conditions that supposedly created creoles under other nations. This presentation will revisit a hypothesis I originally presented in the 1990s, which solves all of these conundrums by placing the origin of plantation creoles at slave castles on the West African coast. The evidence includes documentation from the slave castles themselves. I will also refer to support for the argument that has emerged over the past two decades.