We examine collisions in pulsed laser fields within the framework of multimode Floquet theory; the approach as outlined in a recent paper allows us to deal with collisional problems in pulsed fields without having to explicitly follow the varying field amplitude. Our focus here is on pulses which are on a time scale of the order of the collision time or less; the interest follows from an analysis made by Lee and George suggesting that unusually strong radiative transitions are possible in such fields. We examine transitions in a model problem based on the Na-Ar collision system and conclude, as others have done for this system, that a short, intense pulse is indeed effective in promoting transitions. Our assessment is, however, only partly based upon the traditional measure of examining how transition probabilities vary as the pulse length is shortened/lengthened. This traditional measure, we argue, is not sufficient for investigation of the mechanism of Lee and George because the measure includes varying contributions to the pulse effect as the pulse frequency profile changes - and this is not accounted for in the analysis of Lee and George. We combine, then, the traditional measure with an examination of the effect of changes in pulse-frequency profile. More particularly, results are determined which embody the same frequency profile as the pulse train, but are generated from a series of fixed-amplitude sources and these results are, in turn, compared with ones generated from a single fixed-amplitude source operating at the central frequency of the train. While, as noted above, we conclude for the Na-Ar system that a short, intense pulse is indeed effective in promoting transitions, we find that combining the traditional measure with an investigation of the role of frequency-profile changes significantly alters assessment of the pulse effect.