The measuring uncertainty requirements of digital imaging systems have already become few tens of nanometers in some areas, and expected to be as small as few nanometers in near future. This stringent requirement of uncertainty can hardly be achieved unless the measuring uncertainty is predicted and controlled down to sufficient level at design stage. Digital imaging system comprises many sub-devices such as illumination optics, imaging optics, image acquisition device and image processing routines, and each device has various source of errors. Therefore, to predict the measuring uncertainty at design stage, each source of errors has to be quantified either by theoritically or experimentally, and the mechanism through which these source of errors transfer to result in the measuring uncertainty has to be well understood. This investigation starts with a system modeling, showing how input signal, together with variuos source of errors, propagates from one sub-device to another, and end up as signal output. Then using the system model, the effect of each error source on the measuring uncertainty has been determined. Then finally, the measuring uncertainty of an examplary system has been evaluated and the way of reducing the uncertainty has been introduced.