Two-dimensional nuclear magnetic resonance (NMR) spectroscopy provides more chemical structural information than is provided by 1D NMR spectroscopy. However, the efficient reconstruction of high-resolution spectra is difficult because of the long measurement time and low resolution, which depend on the number of increments in the indirect dimension. We address these problems by applying truncation and non-harmonic analysis (NHA) to the spectra recorded in 2D NMR experiments. The truncation method can improve the signal-to-noise ratio (SNR) and shorten the measurement time, and NHA can reconstruct a high-resolution spectrum irrespective of the window length of the analysis. In this study, the performance of the NHA was verified from both phantom 2D NMR data and experimental data. The data are analyzed by applying NHA and the results compared with those analyzed using Fast Fourier Transform (FFT), a method commonly used for NMR analysis. By truncating the data uniformly, the amount of data required to maintain the spectral peaks is determined. Consequently, we demonstrate that our proposed method can yield accurate spectral information for reductions of up to 6.3% of the original data volume, except for the denser parts. This is expected to improve the efficiency of 2D NMR experiments.