Diesel emissions are a major contributor to combustion-generated airborne ambient particles. To understand the role of diesel particulate emissions on health effects, it is important to predict the actual particulate dose deposited in the human respiratory tract, with respect to number, surface area and mass. This is complicated by the agglomerate nature of some of these particles. In this study the respiratory tract deposition fraction in the size range 10–500 nm, was determined for 10 healthy volunteers during both idling and transient engine running conditions of a heavy duty diesel engine. The aerosol was characterized with respect to both chemical and physical properties including size resolved particle effective density. The dominating part of the emitted particles had an agglomerate structure. For those formed during transient running conditions, the relationship between particle mass and mobility diameter could be described by a power law function. This was not the case during idling, most likely because of volatile compounds condensing on the agglomerates. The respiratory tract particle deposition revealed large intra-subject variability with some subjects receiving a dose that was twice as high as that of others, when exposed to the same particle concentration. Associations were found between total deposited fractions (TDF), and breathing pattern. There was a difference between the idling and transient cycle with TDF being higher with respect to number during idling. The measured size-dependent deposition fraction of the agglomerated exhaust particles from both running conditions was nearly identical and closely resembled that of spherical hydrophobic particles, if plotted as a function of mobility diameter. Thus, for the size range covered, the mobility diameter could well describe the diameter-dependent particle respiratory tract deposition probability, regardless of the agglomeration state of the particles. Whilst mobility diameter well describes the deposition fraction, more information about particle characteristics is needed to convert this to volume equivalent diameter or estimate dose with respect to surface area or mass. A methodology is presented and applied to calculate deposited dose by surface area and mass of agglomerated particles. The methodology may be useful in similar studies estimating dose to the lung, deposition onto cell cultures and in animal studies.