Dual-comb spectroscopy (DCS), with its high resolution and rapid scanning capability, has been widely applied in combustion diagnostics and quantitative gas detection. However, under low signal-to-noise ratio (SNR) conditions, the parameter retrieval accuracy of traditional least-squares-based spectral fitting methods deteriorates significantly. To address this issue, this study proposes a dual-comb spectroscopic grid-search algorithm based on zero-mean normalized cross-correlation (ZNCC). The method constructs a two-dimensional temperature–pressure template library, computes the ZNCC between the equivalent line-averaged test interferogram obtained along a single absorption path and the template signals, and normalizes and interpolates the correlation coefficient matrix using a two-dimensional spline. The optimal temperature and pressure are then determined from the correlation peak position. Simulation results demonstrate that the proposed algorithm achieves relative errors below 0.1% for both temperature and pressure under noise-free conditions, and maintains 1–2% prediction accuracy at an SNR of 15 dB. Compared with conventional least-squares fitting approaches, the ZNCC-based grid-search method exhibits stronger noise immunity and computational stability, enabling single-frame, fast parameter retrieval.