The aim of the current research is to develop a STEM-based digital classroom learning model based on the 21st-century skills and characteristics of elementary school students. To enable the dissemination of this learning model, further research is needed to determine the effect of the STEM-based digital classroom learning model on the 21st century skills of elementary school students. The lack of 21st-century skills among elementary school students and the necessity for a study of the influence of higher-order thinking skills as one of the capabilities that primary school students must develop are contributing factors to this research. This advancement would lead to an understanding of how STEM and HOTS affect student's skills. This was experimental research with 100 fourth-grade students. The data analysis methods employed by SPSS included two-way ANOVA tests, homogeneity tests, normality tests and descriptive tests. The study's findings revealed that: (1) students who are taught with STEM-based digital and conventional learning models have different 21st-century skills, (2) students with high and low higher- order thinking skills have different 21st-century skills, and (3) students with high and low higher- order thinking skills who are taught with STEM-based digital and conventional learning models have a connection to 21st-century skills. Education professionals can use the implications of this research as a guide to develop 21st-century skills for elementary school students.