Interference is a pivotal issue of a non-dispersive infrared (NDIR) sensor and analyzer. Therefore, the main contribution of this study is to introduce a potential method to compensate for the interference of the NDIR analysis. A potential method to compensate for the interference of a nitric oxide (NO) NDIR analyzer was developed. Double bandpass filters (BPFs) with HITRAN (high-resolution transmission molecular absorption database)-based wavelengths were used to create an ultranarrow bandwidth, where there were least-interfering effects with respect to the coal-fired power plant emission gas compositions. Key emission gases from a coal-fired power plant, comprising carbon monoxide (CO), NO, sulfur dioxide (SO2), nitrogen dioxide (NO2), carbon dioxide (CO2), and water (H2O) (in the form of vapor), were used to investigate the gas interference. The mixtures of those gases were also used to investigate the performance of the double BPFs. We found that CO, CO2, SO2, and H2O significantly affected the detection of NO when a commercial, single narrow BPF was used. In contrast, the double BPFs could remove the interference of CO, NO2, SO2, and CO2 in terms of their concentrations. In the case of H2O, the filter performed well until a level of 50% relative humidity at 25 °C. Moreover, the signal-to-noise ratio of the analyzer was approximately 10 when the double BPFs were applied. In addition, the limit of detection of the analyzer with the double BPFs was approximately 4 ppm, whereas that with the commercial one was 1.3 ppm. Therefore, double BPFs could be used for an NO NDIR analyzer instead of a gas filter correlation to improve the selectivity of the analyzer under the condition of a known gas composition, such as a coal-fired power plant. However, the sensitivity of the analyzer would be decreased.