A measurement system has been developed for determining typical long-term speech and background noise levels in university classrooms during lectures. A particular objective was to determine typical levels of student activity noise. A total of 18 lectures were recorded at up to four positions in 14 classrooms. These recordings were digitized and filtered using digital octave-band and A-weighting filters. Software processed the resulting pressure time histories as follows: The signals were squared; short-term mean-square pressures were calculated; sound-pressure levels were calculated; sound-pressure-level frequency distributions were determined and plotted; and the resulting distributions were fitted by one, two, or three normal-distribution curves to identify peaks. Three curves generally gave the best fit. The levels of the peaks so identified were associated with speech, student activity, and background noise levels. The corresponding values were determined for each test case; the results are summarized and conclusions are drawn.