Polyadic family conflicts and the use of alliances were examined in 39 families during naturalistic home observations of mothers (M age = 32.8 years), fathers (M age = 34.6 years), older (M age = 6.3 years) and younger siblings (M age = 4.4 years). The data included transcripts of audio-recorded researcher notes and family member verbalizations for each of the six 90-minute sessions. Using these transcripts, conflict initiators, topics, and resolutions, as well as additional party roles (e.g., alliance, mediator; Black & Baumgartner, 1983) were coded. To distinguish between the amount of support that allies provided to the conflict, alliances were separated into two categories: Cooperative and surrogate alliances. The findings revealed that despite all family members being involved in polyadic conflicts, children tended to be initiators, while parents became involved as additional parties. Alliances occurred more often than the other additional party roles, were likely to be formed when conflicts arose about obnoxious behaviours, and often resulted in the alliance achieving their goals and winning the conflict. Intergenerational alliances (i.e., parent-child) were more likely to be formed than intra-generational alliances (e.g., parental). A preliminary content analysis found that cooperative and surrogate alliances were quantitatively different with respect to the number of turns that allies used when supporting a combatant. Moreover, rule enforcement, control, informational, repetition, and induction were identified as resources that allies used to support their side. Results are discussed relative to literature and theory, with respect to understanding each family member’s social and metacognitive skills and understanding.