Black carbon is a primary aerosol tracer for high-temperature combustion emissions and can be used to characterize the time evolution of its sources. It is correlated with a decrease in public health and contributes to atmospheric warming. Black carbon measurements are usually conducted with absorption filter photometers, which are prone to the filter-loading effect – a saturation of the instrumental response due to the accumulation of the sample in the filter matrix. In this paper, we investigate the hypothesis that this filter-loading effect depends on the optical properties of particles present in the filter matrix, especially on the black carbon particle mixing state. We conducted field campaigns in contrasting environments to determine the influence of source characteristics, particle age and coating on the magnitude of the filter-loading effect. High time resolution measurements of the filter-loading parameter in filter absorption photometers show daily and seasonal variations of the effect. The variation is most pronounced in the near-infrared region, where the black carbon mass concentration is determined. During winter, the filter-loading parameter value increases with the absorption Ångström exponent. It is suggested that this effect is related to the size of the black carbon particle core as the wood burning (with higher values of the absorption Ångström exponent) produces soot particles with larger diameters. A reduction of the filter-loading effect is correlated with the availability of the coating material. As the coating of ambient aerosols is reduced or removed, the filter-loading parameter increases. Coatings composed of ammonium sulfate and secondary organics seem to be responsible for the variation of the loading effect. The potential source contribution function analysis shows that high values of the filter-loading parameter in the infrared are indicative of local pollution, whereas low values of the filter-loading parameter result from ageing and coating during long range transport. Our results show that the filter-loading parameter can be used as proxy for determination of the particle mixing state, thus allowing for differentiation between local/fresh and transported/aged particles.