The present investigation adopts an embodied cognition perspective to characterize students' learning interactions between gesture and haptic experience. Haptic technologies have emerged as a promising approach to science learning because they can simulate physical forces that students would otherwise not experience. Student-generated gestures complement haptic technology by offering insight into students' embodied understandings of physical forces. In this study, we examined the role of tangible haptic molecular models by analyzing students' spontaneous gestures and speech after their learning experiences with these models. We designed three learning conditions—Individual Haptic Model, Aggregate Haptic Models, and Equation-Based Instruction—to prepare undergraduate students for learning from an agent-based computer simulation of dynamic equilibrium. Guided by Knowledge in Pieces (KiP) theory, we employed a think-aloud protocol to investigate how haptic and computational environments cue students' knowledge resources and spontaneous gestures. Our analysis of students' speech and gesture revealed that the haptic experiences—particularly individual-level haptic—augmented the computational environment by fostering shifts in students' reasoning (1) from the phenomenological primitive equilibrium to balance and (2) from representing aggregate patterns to individual interactions. We discuss the implications of using gesture as a window into students' developing reasoning, as well as the role of haptic technology in designing embodied learning environments.