Across southern Scandinavia are some 30 known weapon depositions made in former lakes from the Roman Iron Age (0-375 CE), seven of which are large-scale war booty sacrifices entailing material culture from whole defeated armies. The major weapon deposits contain spears, lances, shields, swords, knives and other militaria, representing a colossal sum of iron – at one site, some 6000 objects totalling 500kg of iron metal. This thesis presents an enquiry into the provenance of the iron used to make these weapons. The aim was to learn more about the origins of the armies through provenancing their weapons. Using laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS), as well as other analytical techniques, this study analysed 13 lances, a single spear and 10 knives from two of the main weapon deposits, and was able to demonstrate regional and subregional provenance hypotheses for each item. The results demonstrate that the iron used in weapon manufacture originated from across Scandinavia, notably Norway and Jutland. The sampled weapons were studied metallographically, showing the lances to be highly standardised weapon products made in a single workshop by distinguished crafstpeople. The knives reflect a diverse range of construction methods associated with localised non-specialist manufacture. The results were used to support an existing 'band of brothers' theory, instigating that the armies were made up of individuals from across Scandinavia (represented by their personal items) who were issued standardised war gear as part of a larger collective force. In order to provenance the iron weapons, it was necessary to develop a robust analytical method. This involved studying the behaviour of trace elements in experimental smelting systems to improve the existing methodology by further validation and refinement. The results from the method development are deemed to be a significant step in iron provenance studies in archaeology.