Three temperature protocols with different peak inert mode temperature (Tpeak-inert) were compared based on source and ambient samples (both untreated and extracted using a mixture of hexane, methylene chloride, and acetone) collected in Beijing, China. The ratio of EC580 (elemental carbon measured by the protocol with a Tpeak-inert of 580 °C; similar hereinafter) to EC850 could be as high as 4.8 for biomass smoke samples whereas the ratio was about 1.0 for diesel and gasoline exhaust samples. The EC580 to EC850 ratio averaged 1.95 ± 0.89 and 1.13 ± 0.20 for the untreated and extracted ambient samples, whereas the EC580 to EC650 ratio of ambient samples was 1.22 ± 0.10 and 1.20 ± 0.12 before and after extraction. It was suggested that there are two competing mechanisms for the effects of Tpeak-inert on the EC results such that when Tpeak-inert is increased, one mechanism tends to decrease EC by increasing the amount of charring whereas the other tends to increase EC through promoting more charring to evolve before native EC. Results from this study showed that EC does not always decrease when increasing the peak inert mode temperature. Moreover, reducing the charring amount could improve the protocols agreement on EC measurements, whereas temperature protocol would not influence the EC results if no charring is formed. This study also demonstrated the benefits of allowing for the OC and EC split occurring in the inert mode when a high Tpeak-inert is used (e.g., 850 °C).