Methods used in diesel emissions measurement at the Bartlesville Petroleum Research Center are described; limitations, adequacy, and needs for further development of each are discussed. Smoke measurements are reported from work with the Hartridge meter, as well as newly developed instruments that are used to view smoke plumes directly, and which seem to offer advantage over smokemeters previously used. Experience in odor assessment by a human panel using reference odor materials is reported as encouraging. Odor intensity is judged with much greater reliability than odor quality; capability to assess the latter remains wholly inadequate.Results in application of the methods for measuring diesel emissions are intended to illustrate the use of experimental techniques to reveal engine and fuel factors as they influence the character, amount, and air-polluting effect of diesel emissions. Although the data do not permit reliable assessment of any of the factors, they are useful in showing orders of magnitude and possible relative significance of the respective emissions in the several categories.