Research over the last four decades has shown that the classroom learning environment impacts on students' cognitive and affective outcomes. Different approaches have been taken to measure students' perceptions of their learning environment, and this has led to the development of a large number of survey instruments. One such instrument is the College and University Classroom Environment Inventory (CUCEI) which was developed in 1987. The CUCEI was recently modified and both the original and modified versions were validated in a number of studies. This article reports the use of the modified CUCEI in two independent studies in computing classrooms in secondary schools and tertiary institutions in Wellington, New Zealand. The statistical performance of the CUCEI was not completely satisfactory in either study and revealed a number of problems common to these two studies. The difficulties are discussed and recommendations made to increase the validity and reliability of this instrument.