Andersen samplers were invented to study the health effects of airborne particles according to the particle size and have been carefully validated and widely used. Recently, it turned to be necessary to quantify the cell concentration of airborne microbes with the progress in understanding the roles that bioaerosols play in the atmosphere. To collect enough particles on filters, collection time was frequently extended from minutes to hours and even days. However, prolonged impaction time may cause considerable microbes trapped in upper stage filters to drop to subsequent stage filters, leading to supposed uncertainties in the microbial size distribution. This study investigated the uncertainties in bacterial cell number size distribution measured with 8-stage Andersen samplers at a flow rate of 28.3 L min−1 (50% cutoff diameters: >11, 7.0, 4.7, 3.3, 2.1, 1.1, 0.65 and 0.43 μm). Results showed that the concentration of bacterial cells in the size range of >4.7 μm could be underestimated by 40–50% as the concentration in the size range smaller than 3.3 μm was overestimated when the sample collection time was more than 6 h. Sample collection time should be less than 20 min to suppress the uncertainty below 10%, and 42 min to be smaller than 20%. Based on identified exponential inverse relation between the dropping rate from each stage and sample collection time, a scheme is proposed to calibrate the counting results of Andersen sampler samples to obtain the number size distribution of airborne bacterial cells with examples.