Measurements of peroxy radical concentrations (2>+2>) were made by the ''chemical amplifier'' technique during the four intensives of the Mauna Loa Observatory Photochemistry Experiment 2 (MLOPEX 2) at the Mauna Loa Observatory in 1991--1992. In this study these data are compared with the theoretical values of the peroxy radical concentrations obtained from steady state analysis of the complete suite of trace gas measurements and other relevant parameters also measured during the experiment. The data from 33 days of the study contain time overlap of the concentration and physical data which allow a meaningful theoretical treatment. The experimental results for 2>+2> agree well with theory for many of the days, but are significantly suppressed from the theoretical expectations on other days. Two hypotheses are presented and tested to explain the observed suppression. The first involves the reaction of the peroxy radicals at aerosol surfaces. The second proposes the loss of through its reaction with unknown and undetected species to develop peroxy radicals subsequently to which the ''chemical amplifier'' is insensitive. Evidence at hand does not allow a clear choice between these or possible alternative explanations. The data suggest that the net rate of O3 generation in the free troposphere is about -1.5 parts per billion by volume per day (ppbv d-1; 24-hour average). ¿ American Geophysical Union 1996