Our laboratory previously developed a method for assessing experimentally induced pain perception through a 2-min constant heat pain stimulation. However, the traditional analysis relying on group means struggles to interpret the considerable inter-individual variability due to the dynamic nature of the response. Recently, trajectory analysis techniques based on extended mixed models have emerged, providing insights into distinct response profiles. Notably, these methods have never been applied to pain paradigms before. Furthermore, various socio-demographic and neurobiological factors, including endocannabinoids, may account for these inter-individual differences. This study aims to apply the novel analysis to dynamic pain responses and investigate variations in response profiles concerning socio-demographic, psychological, and blood endocannabinoid concentrations. 346 pain-free participants were enrolled in a psychophysical test involving a continuous painful heat stimulation lasting for 2 min at a moderate intensity. Pain perception was continuously recorded using a computerized visual scale. Dynamic pain response analyses were conducted using the innovative extended mixed model approach. In contrast to the traditional group-mean analysis, the extended mixed model revealed three pain response trajectories. Trajectory 1 is characterized by a delay peak pain. Trajectory 2 is equivalent to the classic approach (peak pain follow by a constant and moderate increase of pain perception). Trajectory 3 is characterized by extreme responses (steep peak pain, decrease, and increase of pain perception), Furthermore, age and blood anandamide levels exhibited significant variations among these three trajectories. Using an innovative statistical approach, we found that a large proportion of our sample had a response significantly different from the average expected response. Endocannabinoid system seems to play a role in pain response profile.