The dispersion and chemical evolution of NOx in ship plumes has been investigated for marine boundary layer (MBL) conditions. This effort has involved combining a plume dispersion algorithm with a time-dependent photochemical box model. The analysis has considered several factors, all of which can influence the net impact of NOx on the background environment. These include the following: season of the year, latitude of point of release, meteorological setting, and ship NOx emission rate. Reaction rates within a plume were shown to be a nonlinear function of the levels of NOx, leading to relative estimates of ship plume NOx lifetimes that were factors of 2.5--10 times shorter than for ambient marine conditions. The shortened ship-plume NOx lifetime reflects both elevated daytime levels of OH and nighttime levels of NO3 and N2O5, all of which were estimated to be several times larger than those typical of ambient marine conditions. During daylight hours, elevated ship plume OH resulted in the net photochemical production of O3, with peak concentrations being 5--65% higher than background values, depending on latitude. The areal integrated O3 effect, however, is estimated to be quite small due to further plume dilution. In addition, because of the shorter estimated lifetime for NOx, it would seem reasonable that the integrated O3 production from the current Lagrangian modeling effort would be significantly lower than that predicted by global 3-D grid models. The current predicted shortened lifetime for NOx is quite significant in terms of assessing a ship plume's impact on background marine levels of NOx. In fact, these results would seem to explain a significant fraction of the overprediction of NOx levels in and near shipping lanes recently estimated using 3-D Eulerian global models.