The successful integration of robots into various industrial settings depends on optimal human-robot collaboration, where human operators engage the robot to its full capacity and for the most suitable tasks. This optimal collaboration also depends on the operator’s attitude towards robot collaboration. To address this, we developed the Attitude Towards Robot Collaboration (ARC) Scale, which measures the cognitive, affective, and behavioral components of attitude, positioning the scale close to actual collaborative behavior. Item development combined insights from existing robot attitude scales, broader technology-related measures, and practitioner input about human-robot collaboration behaviors. Data were collected from N = 241 participants with prior experience working with robots, and analyzed using Rasch modeling, leading to a selection of 15 items for the scale. The ARC Scale demonstrated good empirical reliability of .77 and an item Infit range between 0.73 and 1.21, the Rasch model showed acceptable fit (RMSEA = .079, TLI = .923, CFI = .924). Construct validity was supported through convergent correlations with the Attitudes Toward Robots Measure, reflecting overlap with general attitudes toward robots, and divergent correlations with the Affinity for Technology Interaction scale, indicating partial differentiation from general technology affinity. Our scale’s focus on collaboration and its inclusion of behavioral components make it particularly suitable as a basis for optimizing human-robot collaboration at the workplace.