To assess the effectiveness of our interactive classroom simulation of the effects of recruitment limitation on student understanding of processes that regulate community structure, we collected data on student perceptions and actual learning during 3 semesters of our Ecology and Evolutionary Biology course at Hope College. Most students in all semesters indicated that they paid greater attention in class (35-55%) when the instructor used a hands-on demonstration, and they reported that they learn concepts better when the instructor uses a hands-on demonstration such as this one (50-55%). However, gains in student learning (measured by comparing pre- and post-demonstration responses to a question about how limited dispersal of offspring from where they’re “born” is likely to affect diversity among species that compete for space) were typically 10% or less, and rarely statistically significant. Whereas more students switched from incorrect to correct responses (14-21%) than from correct to incorrect ones (4-7%), 25-27% responded incorrectly both before and after the demonstration. Direction of switching was not correlated with measures of overall performance in the course. Longer-term retention of gains in understanding was also minimal, as indicated by responses to the same questions on final exams. Other studies by ecology educators have reported similar incongruities between self-reported student perceptions of their own learning and actual tests of that learning. We will discuss the importance of weighing the benefits of classroom demonstrations (as measured by actual learning gains) against the costs in terms of instructor time, other concepts not covered, etc.