In recent years, issues related to sustainable development become more complex and require future engineers to be prepared to face challenges ahead. Therefore this study was carried out to development sustainable behavior instrument to assess engineering students' sustainable behavior and identify whether demographic factors, internal and external factors led to such behavior. 379 respondents were randomly selected in three location of the technical university in Malaysia which are UTHM, UniMAP and UMP. The instrument used was a set of questionnaire consist 43 items on sustainable behaviour of engineering students. Based on the Rasch model analysis using the Winstep software, the items were selected based on the values of reliability, separation index, Point Measure Correlation (PTMEA CORR), Mean Square (MNSQ) and Z Standard (Zstd). The value for Alpha Cronbach (0.94), individual reliability (0.91) and item reliability (0.98) are greater than 0.80 is consider as good and high reliability. As for the items polarity analysis, the data showed that all items are positive. Meanwhile Point Measure Correlation (PTMEA CORR), Mean Square (MNSQ) and Z Standard (Zstd) were used to determine infit items for each attribute. Consequently, four items of sustainable behaviour were excluded; B227, B232, B233, B234. Most of the item suggested to be dropped are related attitudes towards the use of the vehicle. However, statistically analysis indicate the rest of the items measured the sustainable behaviour construct. This study are able to test items validation in the sustainable behaviour instrument for engineering students in higher learning institutions.
