Peer assessment is an evaluation in which the students are assessing each other. However, do they understand what to assess, how to assess, and the objectivity of the assessment? This study aims to analyse students' ability to conduct peer assessments. The research applied a quantitative research design using questionnaire. The participant was a pre-service teacher who assessed their peers' teaching practices focused on basic teaching skills using questionnaire. The data of ten skills were obtained and analysed using a multi-rater Rasch model with Many-Facet Rasch Measurement. Findings revealed that students' assessment was reliable (0.99), with a high separation index (9.46). The chi-square test showed a significant difference between the levels of the assesses. The exact agreement score was 47.3%, and the expected score was 46.8%. The data showed that the assessors did the scoring independently, the distribution of respondents and items' difficulty levels with the same scales were clear. In conclusion, the results indicated that the students were able to conduct peer assessment objectively. These findings provide support for using peer assessment as an objective and an authentic assessment. For recommendation, the grading rubric should be arranged as clear as possible, conduct assessor training, and more often tests to reduce bias.