Introduction: Conflicts occur occasionally in every close relationship and they are inevitable. But the way people communicate during these conflicts is predictive for relationship quality, satisfaction and relationship dissolution. The usual ways in which one communicates during relationships conflicts is called a communication pattern. Some patterns help resolve conflicts, while other patterns intensify conflicts. The constructive communication pattern involves both partners equally in resolving the conflict or stress within the relationship. It refers to a pattern whereby one couple member engages in joint problem solving and mutual disclosure during a conflict situation with the other couple member. Since constructive communication is an intentional behavior, it is plausible that the Theory of planned behavior (TPB) model (Ajzen & Fishbein, 1977) predicts such behavior. According to the TPB, behaviors can be predicted by behavioral intent which is formed based on attitudes toward the specific behavior, the social norm associated with the behavior and the perceived behavioral control to perform the behavior. Since attitudes, social norm and perceived behavioral control are guided by other cognitive and motivational processes, it is expected that one's time perspective and attributional style should help in understanding relationship communication. Thus, the main goal of this study was to investigate whether constructive communication during conflicts can be predicted using the TPB and to determine if expanding the model by including time perspectives and attributional style improves this prediction. Methodology: The sample consisted of 240 young adults, aged between 18 and 30, who were in a serious relationship but not married. The research was conducted in two time points. In the first, elements of the TPP (attitudes towards constructive communication, social norm for such behavior, and perceived behavioral control over constructive communication during conflicts) were assessed using a questionnaire constructed according to Ajzen's (2011) guidelines specifically for this study. Time perspectives (Zimbardo's Time Perspective Inventory, ZTPI, Zimbardo & Boyd, 1999), attributional style (Attributional Style Questionnaire, ASQ, Peterson, Semmel, Von Baeyer, Abramson, Metalsky, & Seligman, 1982) and attributions of partner's negative behavior (Relationship Attribution Measure, RAM, Fincham & Bradbury, 1992) were also assessed. A month later, actual constructive communication during conflicts in the past month were measured using the Communication Patterns Questionnaire's (CPQ) Constructive Communication subscale (Christensen & Sullaway, 1984). Using structural equation modelling, two models were tested: the basic TPB model, and the extended TPB model which included time perspectives, attributional style and attributions of partner's negative behavior. Results and discussion: The results revealed that constructive communication during conflicts can be predicted using the TPB's model. A more positive attitude towards constructive communication, a stronger social norm for communicating constructively and higher perceived behavioral norm were associated with a stronger intent to communicate constructively in future conflicts. Behavioral intent predicted actual constructive communication during conflicts. Attitudes, social norm, and perceived behavioral control did not have a direct effect on constructive communication, but rather an indirect one through behavioral intent. Examining the extended model, in which time perspectives, attributional style and attributions of partner's negative behavior were included in the TPB model, it was shown that the extended model improved the prediction of constructive communication compared to the basic TPB model. Past positive and negative time perspectives predicted subjective norm, and present fatalistic time perspective was associated directly with lower behavioral intent to communicate constructively. Those more oriented towards the past showed a stronger norm for constructive communication, although those focused on negative aspects of their past believed they were not able to communicate constructively. Lower behavioral intent was predicted by present fatalistic time perspective, which is in line with previous findings indicating present fatalistic oriented individuals generally tend to underestimated their behavioral intent in self –report measures (Van Ittersum, 2012). Perceived behavioral control was also associated with attributional style's stability dimension. The stability dimension of attributions of partner's negative behavior had a direct effect on constructive communication, as well as an indirect one through attitudes towards constructive communication. Assigning blame to the partner for what he did (dimension of blame in the attributions of partner's negative behavior) had only an indirect effect on constructive communication through perceived behavioral control. Therefore, we can say that those who believe their partner's negative behavior was caused by a one-time only cause as well as those who do not blame their partner for what he did are more likely to communicate constructively. Conclusion: The present research contributes to better understanding of intimate relationship conflicts and communication, as well as their antecedents. A deeper understanding of the cognitive and motivational basis of constructive communication, through attitudes, social norm, perceived behavioral control, behavioral intent and factors influencing them provides one of the preconditions for the integration of these constructs into a model that not only broadens the theoretical knowledge of the field, but is also applicable in everyday life couple's therapy and interventions. Indirectly, the research fuels further advancement in the study providing interesting directions for future research.