This research aimed to analyze the levels of agreement between kindergarten teachers and principals in identifying the assessment of teachers' teaching competencies and performance. The study was designed under a quantitative approach using a survey. It implemented a non-probability sampling technique with purposive sampling. The sample of the population comprised of 173 kindergarten teachers and 101 principals in Semarang District, Indonesia, or a total of 274 respondents. The data were collected through a questionnaire and analyzed using Cohen's Kappa coefficient to measure the levels of agreement between raters and Pearson Chi-Square test was also utilized to determine the differences in perceptions among principals and teachers. The findings showed that the levels of agreement between raters were averagely in the no agreement category, implying the existence of differences in perceptions among teachers and principals. The involvement of a multi-rater strategy in such research is a rare effort, especially for the Early Childhood Education (ECE) level in Indonesia. Researches regarding teaching competencies and performance generally only involve single rater, either teachers or principals who judge themselves on their competencies and performance, thus the results tend to be subjective. In conclusion, the assessment of teaching competencies with the relation of cognitive abilities was conducted through a test that considered subjective questions and case analysis to evaluate the teachers' skills based on their performance and self-description. Both personal and social assessments utilized self-assessment forms or autobiographies, which were completed with specific themes. Meanwhile, the performance assessment was observed with the assessment rubric and comparison with the learning process performed by an individual educator.