Accurate characterizations of particulate matter based on reliable measurement techniques are essential to monitor, regulate, and model atmospheric pollution levels, so that a unified scientific database can be achieved to guide progress in air quality and emission policies. Of particular interests are the relevant sizes of combustion-generated particulates that are typically fractal aggregates of small spherical primary particles. In this study, three independent diagnostics-mobility particle sizer, transmission electron microscopy, and laser scattering and extinction were critically compared and evaluated for their abilities and limitations in characterizing properties of polydisperse soot populations emitted from well-defined laboratory flames. For the present experimental conditions, microscopy and optical measurements were found to agree not only with each other but also with past studies. While the mobility sizing technique could not reveal information on spherule diameters and fractal properties due to the spherical particle assumption, it yielded overall equivalent sizes that were comparable to aggregate sizes. The measured mobility diameters, however, differed from the other two techniques for aggregate size distributions and were not consistent with expected sooting behaviors when the flame conditions were changed. The maximum discrepancy between the mobility and optical measurements was a factor of three for particle volume fraction and more than an order of magnitude for specific surface area. These results suggest further experimental and theoretical investigations on the correspondence between mobility diameter and actual properties of aggregated particulate matter.