AbstractHabitat preferences of fifteen species of ungulate were studied between July and September 1975 in dry savanna woodland of Mwabvi Game Reserve, Southern Malawi, Central Africa.Thirteen vegetation types were recognized using vegetation transects and plotless sampling techniques (but three of these are omitted from further analysis here either because of the difficulty of sampling in a representative manner or because of their infrequency). A combination of six methods was used to study habitat preferences since each had particular advantages or disadvantages, although quadrat sampling of fresh pellets/spoor counts was the most satisfactory and so used predominantly.The most common and widely distributed species was the Greater kudu, although abundance estimates in different habitats varied according to the method used. Pellet/spoor counts showed heavy use of open grassland by kudu which was not detected by daytime drives. Temporal patterns of activity indicated that the species was moving into more closed woodland during the day and only coming out into open grassland at night. Relative abundance estimates of a species may also be affected by changing group size. The mean group size of kudu increased in more open vegetation, although group size in warthog did not show this relationship. Some species, e.g. impala, reedbuck, and klipspringer showed narrow habitat preferences, occurring in only one or a few vegetation types, whilst others, e.g. kudu, sable, duiker, suni, and grysbok showed much broader preferences.MixedColophospermum mopane /Acaciawoodland had the highest species diversity but rather low abundance. Open grassland andAcacia/Albiziawoodland had high species diversity together with the greatest abundance of ungulates. OpenJulbernardia/Diplorynchus, Brachystegia/Julber‐nardiawoodland, and particularlyPterocarpus/Diplorynchus/Combretumhill woodland had the lowest species diversity and abundance of ungulates. Ready access to cover and availability of suitable food plants probably account for the preferences of certain vegetation types over others, whilst the different behavioural ecology and food preferences of each species account for individual species differences. Access to water probably also determines habitat preferences in particular species, and studies during the wet season may show different distribution