Caking of food powders is highly undesirable. Many food powders are powder ingredient mixes and there is little work reported on the caking of food powder mixes. This study applied a combination of cake strength and moisture content measurements, coupled with visual assessment of the bulk powder and two particles in contact to investigate the humidity caking behaviour of the powders over time. The food powder binary mixes consisted of a 40% or 80% "sticky" powder (whey permeate WP or maltodextrin MD) and a "non-sticky" powder (salt, flour or paprika). The powders were exposed to 76% relative humidity to make the WP and MD sticky. The WP mixes had weaker but similar cake strength profiles to the WP powder. For the salt mixes with both WP and MD, there was a continuous uptake of water over time resulting in very weak salt mixes. This may be caused by lowering of the deliquescence relative humidity of salt when contacted with WP or MD. The caking behaviour of the MD mixes depended on the non-sticky powder. The MD/flour mixes had a weaker but similar profile to the MD powder, while MD/paprika and MD/WP displayed a maximum cake strength after 6-10 days.