The Lokken orebody is situated within a 1- to 2-km-thick Lower Ordovician metavolcanic sequence of ophiolitic affinity. The deposit is probably the largest ophiolite-hosted massive sulfide body in the world and originally contained 25 to 30 million metric tons of ore. It has a primary elongated morphology; the shape was largely controlled by its sheetlike, fissure-related, hydrothermal feeder zone which underlies the central part of the stratiform sulfide body along its entire length. The feeder zone is very well defined down to at least 300 m below the contemporaneous paleosea floor.The main part of the feeder zone, with a width of approximately 100 m, comprises sulfide vein stockwork and disseminations. It increases in density toward the most central 40 to 50 m and particularly toward the massive sulfide ore level. Host volcanites have suffered pervasive Fe-rich chlorite + quartz alteration. Close to the stratiform ore level, sericite gradually becomes more prominent at the expense of chlorite.Outside the main zone, the sulfide content decreases rapidly, with more moderate alteration of host volcanites to chlorite + albite + or - epidote assemblages, containing Fe or Mg-rich chlorites. The general mineralization paragenesis and interpretation of the evolution of the feeder zone is (1) early precipitation, in a zone which is at least 130 m wide, of fine-grained dirty pyrite and sphalerite in thin veins and as void fillings and disseminations, resulting from mixing of ascending hydrothermal solutions and seawater--later alteration of wall rocks produced Mg-rich chlorite and albite; (2) partial sealing of the upflow zone due to precipitation of early sulfides, followed by overgrowth of coarse-grained pyrite on the early formed sulfides in a central zone approximately 100 m wide; (3) replacement of earlier sulfides by chalcopyrite and dissolution of most of the sphalerite in the central zone, Fe-rich chlorite-quartz alteration by primary or near-primary hydrothermal solutions; (4) selective dissolution of chalcopyrite and formation of residual pyrite veins in the most central part of the feeder zone; and (5) precipitation of chalcopyrite, SiO 2 , and subordinate sphalerite in vein and fracture fillings during the declining stages of hydrothermal activity.The consistent Mg enrichment of altered rocks suggests either that the primary hydrothermal solutions were Mg bearing or that they had gained Mg by influx of normal ambient seawater from deeper in the upflow zone. Furthermore, data show that, at some stage, the hydrothermal solutions were undersaturated with respect to silica as well as Zn, Cu, and Fe.In parts of the Lokken hydrothermal system, the amount of sulfides precipitated within the feeder zone is more than four times that in the immediately overlying stratiform ore. Isolated feeder-type mineralization of the uppermost gabbro through to the lower part of the pillow lava sequence is considered to represent a nearly total loss of the metal load of hydrothermal solutions during their ascent through the sea-floor crust. It is suggested that precipitation within hydrothermal upflow zones, either by cooling or influx of ambient seawater, may have exerted a significant control on the composition and size of the associated massive sulfide deposit.