This study aimed to (1) identify the characteristics of Scratch-assisted interactive learning media based on problem-solving principles that are valid, practical, and effective, and (2) examine students’ learning trajectories that describe their cognitive processes during learning activities using the developed media. This research employed a Research and Development (R&D) approach using the Plomp development model, which consists of five stages: preliminary investigation, design, realization/construction, testing, evaluation and revision, and implementation. The participants of this study were 15 fifth-grade students from SDTK Pelita Hati. Data were collected through expert validation sheets, teacher and student response questionnaires, mathematical critical thinking tests, and Hypothetical Learning Trajectory (HLT) analysis. The data were analyzed using both descriptive quantitative and qualitative methods. The results indicated that: (1) the Scratch-assisted problem-solving-based interactive learning media was highly valid, with an average validation score of 4.5 from both material and media experts; (2) the developed media was highly practical, as shown by mean response scores of 4.5 from teachers and 4.52 from students; (3) the learning media was effective in improving students’ mathematical critical thinking skills, as evidenced by an increase in the mean score from 39.67 on the pretest to 78.67 on the posttest, with an N-Gain value of 0.65 categorized as moderate improvement; and (4) students’ learning trajectories demonstrated progressive development in mathematical thinking, moving through stages of problem identification, strategy formulation, problem-solving execution, and critical conclusion drawing supported by Scratch-based learning activities. In conclusion, the Scratch-assisted interactive learning media based on problem-solving is feasible and effective for enhancing fifth-grade students’ mathematical critical thinking skills.