Although children are known to be highly sensitive to interparental conflict, important questions remain regarding which specific combinations of positive and negative behaviors as well as verbal and nonverbal expressions are most predictive of children's perceptions. In this pilot study, we examined observational data on interparental conflict as predictors of children's reports of perceived threat and insecurity in 43 families. Fathers' nonverbal negativity was strongly linked to children's perceived threat and insecure family representations, but both parents' nonverbal and mothers' verbal positivity buffered its impact on children. Our findings support previous research findings that parents' negativity may have less adverse effects on children when it takes place in a positive family climate. Interparental conflict is well established as a major predictor of child maladjustment (Cummings & Davies, 2010). Two leading models propose key variables explaining this association. First, the cognitive-contextual model (Grych & Fincham, 1990) posits that children who perceive high levels of threat when faced with interparental conflict are likely to be particularly distressed. Second, emotional security theory (Davies & Cummings, 1994) holds that maintaining felt security is a primary goal for children in the family setting.According to this theory, children from high-conflict homes are expected to develop insecure family representations, marked by low confidence in parents' abilities to manage difficulties for the purpose of preserving family stability. Both dimensions, perceived threat and insecure family representations, were empirically supported in a number of longitudinal tests as explanatory mechanisms linking interparental conflict and children's long-term adjustment (e.g., Cummings, George, McCoy, & Davies, 2012;Grych, Harold, & Miles, 2003). The present study builds directly on these two models, using observational data on interparental conflict to specify the dimensions that predict children's perceptions of threat and insecurity in this context.Previous research has reported that it is not whether couples argue but how they do that is most pertinent to the well-being of children. emerged as a more powerful predictor of children's maladjustment than parents' overt hostility. In line with the notion that children are exceptionally sensitive to interparental tensions, children distinguish "mixed message resolution" (inconsistent in content and emotion, e.g., an angry apology) from consistently positive conflict endings when responding to analogue unresolved conflicts (Shifflett-Simpson & Cummings, 1996). They are also sensitive to adults' emotional tone in their tendency to display negative reactions when faced with interparental conflict (Cummings, Goeke-Morey, Papp, & Dukewich, 2002). In sum, these prior findings showed that (1) observational data of interparental conflict are a particularly reliable measure in the field and (2) the impact of interparental conflict on children is probably best u...