Student test score percentages in the Physiology and Disease (PAD) course at the Illinois Mathematics and Science Academy, a high school for students of the state of Illinois gifted in math and science, were studied over a period of 5 yr. Inquiry-based laboratory experiences in the course were slowly converted during this time from partly student centered and mostly teacher led to completely student centered beginning in fall 2008. Quarterly analysis of the effect of increased inquiry upon average weekly report submissions of 400 students over 4 yr showed a significant improvement in submission (P less than 0.0002) between "quarters 1" and "2" and also improvement from year to year between the academic years of 2006/2007 and 2009/2010 (P less than 0.0001). A comparison of student test score percentages from 346 students in 4 major tests showed a significant increase (P = 0.0125) beginning in the academic year of 2008/2009, when the conversion of all laboratories in the course from partly student centered to completely student centered was concluded compared with scores over the 2 yr from 2006/2007 up to this point. There was also a significant difference (P less than 00001) in test score percentages between the individual tests themselves over the 4 yr studied. Taking the study a step further, the 35 students registered in the two PAD classes offered in the fall 2010 semester were divided in each of their classes into student-centered and teacher-centered groups, with the former designing all their experiments and the latter following instructions from the teacher. Student score percentages on specific test questions from the four major tests that focused on transfer of student understanding were compared between these two groups. There was a significant improvement (P = 0.012) when students designed their own laboratories (student-centered group) compared with doing what the teacher asked (teacher-centered group). There was also a significant difference between these student score percentages among the individual tests (P less than 0.0001). These data suggest that an increase in student-centered experiments may lead to a corresponding increase in test performance on questions involving student transfer. (Contains 7 tables and 3 figures.)