Ion Adsorption Deposits (IAD) are a major source of Heavy Rare Earth Elements (Gd – Lu; HREE) to global markets. Most currently known IAD are hosted in thick regolith profiles, developed in temperate to tropical climates from metaluminous to peraluminous granites that have experienced varying degrees of metasomatic alteration. However, typically these granites are not strongly Rare Earth Element (REE)-enriched. In contrast, alkaline to peralkaline syenitic to granitic complexes are known to contain elevated concentrations of REE – leading to the hypothesis that if IAD can form on relatively REE poor protoliths, then regolith profiles developed upon REE-enriched alkaline to peralkaline rocks should contain IAD more strongly enriched in REE. The alkaline to peralkaline Ambohimirahavavy Complex is covered by regolith profiles developed in a tropical climate. The petrogenesis and controls on REE mineralisation of the Ambohimirahavavy Complex’s south-eastern Ampasibitika intrusion have been investigated here. The Ampasibitika Intrusion is a heterogeneous sub-volcanic intrusion, where silica-undersaturated to –oversaturated intrusive rocks are exposed alongside volcanic units. Petrographic features, mineral chemistry and whole rock major and trace element geochemistry indicate the Ampasibitika Intrusion evolved via fractional crystallisation and differentiation of a weakly peralkaline melt, sourced from low degree partial melting of a metasomatised mantle source, in a shallow crustal magma chamber to emplacement of heterogeneous crustal mushes and melts to higher crustal levels during caldera forming events. The coalescence of highly evolved residual/immiscible melts in the magma chamber roof zone and assimilation of crustal material and/or previously crystallised syenites formed silica-oversaturated trachytic to peralkaline granitic melts, which were emplaced as a concentric dyke swarm. Varying degrees of fractional crystallisation and melt differentiation, and possible salt-silicate melt immiscibility, resulted in variable REE-enrichment. REE concentrations generally increase with progressive melt evolution and volatile-enrichment. The emplacement of crystal mushes and melts to higher crustal levels was associated with melt/fluid immiscibility/exsolution, causing late magmatic to hydrothermal alteration. Mineral stable isotope signatures indicate most alteration fluids were of magmatic origin. Potential source minerals of REE for the Ampasibitika IAD include REE-fluorcarbonate, britholite, allanite, titanite, which are generally late magmatic to hydrothermal mineral phases, and early magmatic calcic-amphibole and -clinopyroxene. However, late magmatic to hydrothermal alteration may also produce REE-poor phases (e.g. sodic-amphibole/clinopyroxene) or REE-enriched minerals resistant to dissolution during weathering (e.g. zircon). Thus, the alteration style controls the development of REE source minerals of varying benefit to IAD formation. Within the Ampasibitika IAD, REE are principally adsorbed to kaolinite and/or halloysite; thus, the formation conditions of kaolinite derived from a range of regolith profiles and depths were investigated using stable isotope methods. The results suggest that, although some kaolinite present is of hypogene origin, these potential REE-host minerals are principally of supergene origin, and therefore the products of intense chemical weathering. Overall, REE mobility during late magmatic to hydrothermal alteration has variably influenced the availability of REE for subsequent IAD. Thus, although magmatic processes may produce extreme REE-enrichment in alkaline to peralkaline intrusions, it is generally the late magmatic to hydrothermal evolution that affects the development of IAD REE-source minerals of the Ampasibitika Intrusion.