A parallel plate differential mobility analyzer (DMA) having 100 independent current collectors is calibrated to relate the axial distances L n between the inlet slit and the detector position to the particle mobility Z at given voltage difference V and sheath gas flow rate Q. Calibrating species are tetraheptylammonium bromide clusters (THABr) and polyethylene glycol (PEG35k, 5 nm in diameter), generated by a bipolar electrospray source, and purified in a cylindrical DMA. Gaussian fitting of the raw discrete mobility spectra in the form of ion current I n versus collector position L n , I n (L n ), yield the mean value L o of the collector position maximizing the signal for a given ion. The many (Z,V,L o ) triads obtained at given Q from many different DMA voltages and standard mobilities collapse into a single 1/(Z i V j ) vs L o curve when slight adjustments are made to the Z i . For different flow rates, Q/(Z i V j ) vs. L o curves collapse also, as long as the peaks are moderately narrow. However, for sufficiently small Q/Z, the THABr cluster peaks become broad, and the curves Q/(Z i V j ) vs. L o cease to collapse precisely. In contrast, the data for PEG show that this behavior is not a low-Q (Reynolds number) effect from the growth of the two lateral boundary layers, but is rather due to the broad and non-Gaussian peak shapes obtained at low Q or high Z. The calibration is accordingly unaffected by the Reynolds number. This simplicity was unexpected, given the three-dimensional flow in this DMA with growing lateral boundary layers.