This descriptive study investigates the integration of peer and teacher assessment in evaluating oral presentations among English as a Foreign Language (EFL) undergraduate students. The research followed a mixed-methods approach aiming to a) analyze the characteristics of peer feedback comments, and b) examine the alignment between peer and teacher scores to assess grading validity and reliability. Data were collected from 50 Spanish students over three academic years (2020-2021, 2023-2024, and 2024-2025), comprising 341 peer feedback forms that included both Likert-scale scores and open-ended comments focusing on four criteria: language use, task achievement, fluency and pronunciation, and communicative resources. Quantitative analyses were conducted to calculate means, standard deviations, Pearson correlation coefficients, and Cronbach’s alpha values. Results indicated strong alignment between peer and teacher assessments in 2020-21 and 2024-25, but a moderate correlation and lower reliability in 2023-24, suggesting inflated peer scores. A qualitative exploration through content analysis and categorization revealed consistent feedback themes across years, with fluency, vocabulary, and pronunciation being the most frequently mentioned strengths and weaknesses. These findings suggest that, when supported by training and structured rubrics, peer assessment can be a valid and reliable complement to teacher assessment in EFL oral presentation grading.