Quantitative laser-based imaging of NO concentrations is important for many practical high-pressure combustion applications. With increasing pressure, however, it is increasingly affected by interference, laser- and signal absorption and temperature variations. In steady laminar flames multiple sequential laser-based measurements can provide corrections to increase the over-all accuracy; however, in practical combustors with turbulent flames not all required correction data can be obtained simultaneously on a single-shot basis. We quantitatively investigate the influence of various corrections on NO laser-induced fluorescence using detailed measurements in laminar methane/air flat flames at 1 - 60 bar. We investigate the influence of O2 interference, the dependence on local temperature and the effect of laser and signal attenuation. Despite using a NO detection scheme with minimum O2-LIF contribution, the fluorescence interference yields errors of up to 25% in the slightly lean 60bar flame. The over-all dependence of NO number density on temperature in the relevant range is low (< 6% for a 200 K temperature variation) because different effects cancel. In contrast, attenuation of laser and signal light by combustion products CO2 and H2O, which is usually neglected, yields errors up to 40% despite the small scale (8 mm diameter) of our experiment. In practical devices, these attenuation effects may be the major source of errors. Understanding the dynamic range for each of these corrections provides guidance to the uncertainty in single shot images at high pressure.