This paper presents our efforts to develop a scale for measuring students' learning experiences with knowledge-based chatbots in massive open online courses (MOOCs) through three studies. In Study 1, we conducted a qualitative synthesis of the current literature and analyzed students' open-ended responses regarding their experiences with a knowledge-based chatbot. Consequently, we identified eight salient domains (i.e., social presence, teaching presence, cognitive presence, self-regulation, co-regulation, perceived ease of use, behavioral intention, and enjoyment), resulting in the creation of 53 items. In Study 2, we selected 30 items that received more than 80% agreement from five experts. Finally, in Study 3, we reported the findings of exploratory and confirmatory factor analyses of the final scale based on student responses (N = 237) and presented 22 items across five domains (i.e., social presence, teaching and cognitive presence, self-regulation, perceived ease of use, and behavioral intention). This research contributes to the current literature by providing an instrument to measure students' learning experiences with knowledge-based chatbots in MOOCs, which is presently unavailable. The scale developed in this study could be employed for further research aiming to systematically develop knowledge-based chatbots and investigate the relationships between salient factors influencing students' learning experiences in MOOCs.