Tree shrews of both sexes exhibit marking behavior ("chinning") in response to scent marks made by the urine of fertile male conspecifics. To isolate the effective odor components, the urine was fractionated by liquid-liquid extraction and TLC, and the fractions were tested by bioassay. The results show that chinning is elicited by several lipophilic urine fractions, which are more effective in combination than alone. To characterize the complex scent signal, the lipophilic extracts from urine of the two sexes were analyzed by GC-MS and compared. The GC profile of the males is distinguished by pyrazine compounds not detected in the profile of the females. The profiles of the sexes also differ with regard to several volatile monocarboxylic acids, which are present at higher concentrations in male than in female urine. More than 30 urine components have been identified. Synthetic equivalents of these urine components were bioassayed for effectiveness in eliciting chinning and compared with one another as well as with scent substances not normally present in tree-shrew urine. Strong chinning responses were elicited by (1) certain pyrazine compounds and (2) some monocarboxylic acids, when presented at the high concentrations specific to male urine. Marking behavior is usually not elicited by scent substances not contained in tree-shrew urine or by urine components common to mammals in general. The data so far available indicate that the male-specific scent signal of tree shrews is based less on a single unique component than on the qualitative and quantitative characteristics of a multicomponent mixture. In the bioassay, tree shrews of both sexes respond equally to the male specific substances by chinning. As the scent signal represented by these substances has a different meaning to males and females (rival or potential mate, respectively), chinning probably serves several different functions.