Large batches (800 g) of nanostructured Mg - x wt % MmNi5 (x = 10 - 70) composites were prepared by ball milling elemental Mg with MmNi5 in an attritor for 12 h under hydrogen atmosphere. There was no alloy formation between Mg and MmNi5 within the milling times employed. The grain size of Mg in the composite varied from 28 to 45 nm, increasing with Mg content, whereas that of MmNi5 was constant at 12 nm, irrespective of its concentration in the composite. BET surface area of the milled composites was in the range of 25 to 58 m2 / g, increased with MmNi5 content. The absorption kinetics and capacities of these composites were measured at 100, 200 and 300 {ring operator} C under 30 bar hydrogen pressure on a small sample of about 8 g. The samples absorbed at all the temperatures, the absorption rate increases with increase in MmNi5 content. But the rates decreased marginally with temperature and attained steady state in less than 400 s even at 100 {ring operator} C. However, the hydrogen absorption capacity of the composites followed the Mg content at all the temperatures, with Mg - 10 wt % MmNi5 showing the highest capacity of 5.1 wt%. MmNi5 has a significant effect on the absorption of hydrogen by Mg at temperatures as low as 100 {ring operator} C, even though hydride of MmNi5 was not present in the hydrogenated composite. The nanostructure of Mg together with distribution of MmNi5 on grain surface/grain boundary of Mg appears to have enhanced the absorption, with MmNi5 probably acting as a conduit for hydrogen diffusing into Mg grain. The performance of hydrogen storage device with larger quantities of Mg - MmNi5 composites (350-500 g) was evaluated in the temperature range of 100 - 150 {ring operator} C with supply pressure of 10-30 bar. The absorption rates and quantity of hydrogen absorbed increased with supply pressure and decreased with temperature. The fraction α was more than 70% in compositions with up to 10 - 30 wt % MmNi5, but was much lower at higher MmNi5 content. The absorption behaviour of these composites followed the same trend as in the case of testing on smaller sample size. © 2006 International Association for Hydrogen Energy.