Event-sampling and scans were used for collecting data on male-infant-male triadic interactions, and their effects on member spacing respectively in a group ofMacaca thibetana at Mt. Emei in 1989. The group was partially provisioned by human visitors in seasons other than winter, and could be observed closely. In addition, a stable linear male-hierarchy among five males existed for two years since the end of 1987, providing a good social condition for this topic. The triadic interactions were specific to the birth season, and recognized as three types being on a continuum functionally changing from passive “agonistic buffering” (4.8%) to active spatial cohesion, which resulted in a significant decline of intermale distances. Positive correlations were documented between the triad initiation rate and the number of females in consort with the males in the mating season (MS), and between the triad reception rate and the number of infants in proximity to the males in the MS when maternal care was significantly reduced. Thus the male's mating effort and kin/sexual selection may deeply be involved in the triad of this species. Considering that the two triad-species,M. sylvanus andM. thibetana, had different levels of paternity, but shared similar foraging conditions, and showed similar intensities of male-infant caretaking, the triad was very likely a by-product of male-infant caretaking, which was probably shaped to compensate heavy maternal investment to young offspring in harsh conditions. Accordingly, the long-term arguments about the triad inM. sylvanus can be united to a model of the way in which “male-infant caretaking” hypothesis works ultimately, and “regulating social relations” hypothesis does proximately.