Background: Caring nurse-patient interaction has been linked to improvements in patient health outcomes, empowerment, and security. Aim: To design a predictive model of caring interaction, based on the development of the Nursing Inter- actions in Caring _ Competence Assessment for Nursing Professionals (NIC_CA-Prof) tool. Methods: A cross-sectional psychometric study was carried out. Descriptive data analysis and exploratory factor analysis were performed. To design the predictive model, partial least squares regression analysis was used in consistent mode. Findings: The sample consisted of 544 nurses with a mean age of 45.9 years old (Standard Deviation = 10.7 years). Exploratory factor analysis yielded 19 items and four factors. A predictive model of four linearly related composites was validated: basic nursing care, therapeutic relationship, problem management, and adaptation. Conclusions: The predictive model provides a visual depiction of the linear and temporal development of the phases of caring interaction, showing the effect of one phase on the next.