Mathematical literacy (ML) is a central competence in mathematics education, particularly for prospective teachers. However, most ML assessment instruments target school students, leaving a gap in instruments designed specifically for prospective teachers. This study aimed to develop and empirically evaluate an ML assessment instrument integrating Sustainable Development Goals (SDGs)-related contexts within the Space and Shape domain. Using a research and development design, the instrument was developed through a literature review, expert validation, readability testing, and a pilot test with 29 prospective mathematics teachers who had completed a geometry course. Data were analyzed using Pearson’s product-moment correlation for item validity, Cronbach’s Alpha for internal consistency reliability, and descriptive statistics to examine performance by item and mathematical literacy process. Nine of the ten items demonstrated statistically significant item-total correlations, while one item was excluded from the reliability analysis. Cronbach’s Alpha was 0.813. Participants' mean performance was 74% in formulating mathematical situations, 65% in employing mathematics, and 58% in interpreting mathematical results. The findings provide initial evidence of the instrument's measurement properties and describe variations in prospective teachers' performance. Further testing with larger and more diverse samples is needed to establish broader applicability.