Low emittance (sub-100 nm rad) measurement of electron beams in plasma accelerators has been a challenging issue for a while. Among various measurement schemes, measurements based on single-shot quad-scan using permanent magnetic quadrupoles (PMQs) has been recently reported with emittance as low as ∼200 nm Weingartner (2012Phys. Rev. Spec. Top. Accel. Beams15 111302). However, the accuracy and reliability of this method have not been systematically analyzed. Such analysis is critical for evaluating the potential of sub-100 nm rad emittance measurement using any scheme. In this paper, we analyze the effects of various nonideal physical factors on the accuracy and reliability using the PMQ method. These factors include aberration induced by a high order field, PMQ misalignment and angular fluctuation of incoming beams. Our conclusions are as follows: (i) the aberrations caused by high order fields of PMQs are relatively weak for low emittance measurement as long as the PMQs are properly constructed. A series of PMQs were manufactured and measured at Tsinghua University, and using numerical simulations their high order field effects were found to be negligible . (ii) The largest measurement error of emittance is caused by the angular misalignment between PMQs. For low emittance measurement of ∼100 MeV beams, an angular alignment accuracy of 0.1° is necessary. This requirement can be eased for beams with higher energies. (iii) The transverse position misalignment of PMQs and angular fluctuation of incoming beams only cause a translational and rotational shift of measured signals, respectively, therefore, there is no effect on the measured value of emittance. (iv) The spatial resolution and efficiency of the detection system need to be properly designed to guarantee the accuracy of sub-100 nm rad emittance measurement.