The background to Friedrich Welwitsch's seven year expedition to Angola and the subsequent fate of his herbarium of 10,000 collections are reviewed. In typifying the approximately 1,000 species names based on his collections, it is important to know where the specimens were at the time of publication of the names. In most cases there are no holotypes, and lectotypifications have to be made. It has not been generally realised that Welwitsch lumped together under one number all the collections which he considered to be conspecific, which means that designation of isotypes is open to serious misinterpretation. Data added retroactively to labels by copying from Hiern's Catalogue may be erroneous. The collection has now been distributed to at least 24 different herbaria. Keywords: Angola, BM, LISU, typification, Welwitsch collections.