Near real-time particle size distribution measurements with a scanning electrical mobility spectrometer (SEMS) requires a fast scan DMA operation. The classification characteristics of the DMA when operated with an up-scan voltage can be established from calculations of particle arrival-times at the sample port of the DMA. This approach, referred to as the arrival-time transfer function (ATF) calculation approach, has been shown to accurately calculate DMA mobility transfer functions under up-scan operation for a range of scan times. Fast SEMS operation, however, are further complicated by downstream effects of particle smearing, plumbing delays (offset time) and counting time interval of the particle counter. These effects have thus far been quantified empirically. In this paper, a theoretical framework is proposed to minimize empirical corrections. The effects of particle smearing, plumbing delays (offset time) and counting time interval are incorporated in determining SEMS transfer functions, which are then used to predict the signal response to a known upstream size distribution under different operating conditions. The predicted signals are then validated with experimental data with scan times as fast 5 s. Using the calculated SEMS transfer functions and a regularization-based inversion algorithm, we are able to accurately invert SEMS signals from a scan as fast as 5 s. Our proposed SEMS transfer function calculation approach, thus, suggests that near real-time, accurate size distributions may be possible with SEMS measurements.