Introduction Momentum turns, in which a team starts progressing (positive momentum) or regressing (negative momentum) in relation to the victory, are often reported in sports. Early and recent theories of momentum assumed that it elicits dynamic changes in psychological and behavioral states (Adler, 1981; Gernigon, Briki, & Eykens, 2010). However, previous research on team momentum has only taken snap shot measures of psychological and performance variables after some (manipulated) momentum period. Therefore, it has remained unknown howpsychological and behavioral team states actually change during positive and negative momentum. ObjectivesIn the current study we applied a dynamical approach in order to provide insights into the nature of psychological and behavioral changes during positive and negative team momentum. MethodEight teams of two rowers finished two ergometer races against a virtual opponent. In the positive momentum race the team gradually progressed and in the negative momentum race the team gradually regressed in relation to the victory. Each minute, the team members verbally responded to a collective efficacy and a task cohesion item. In addition, their exerted efforts and interpersonal coordination were continuously measured with force sensors and Nintendo Wii cameras. ResultsWe found negative changes in collective efficacy and task cohesion during negative momentum, which were stronger than the positive changes during positive momentum. Team members’ efforts rapidly decreased during negative momentum, whereas we found an adaptive sequence during positive momentum. Finally, interpersonal coordination was worse during negative momentum than during positive momentum.ConclusionOur results show that, relative to positive team momentum, negative momentum elicits stronger (opposite) psychological changes and accompanies different (less adaptive) behavioral regulation. Such insights could not be detected with snap-­‐shot research designs. We therefore propose that future applications of dynamical research methods could greatly contribute to the study of real-­‐time psychological and behavioral processes in sports.