<p>Reliable process monitoring is critical to the industrial applications of the laser powder bed fusion (L-PBF) additive manufacturing (AM) technique. This study presents an acoustic emissions-based process monitoring framework where a mapping model between manufacturing parameters and acoustic features is developed to enable real-time health monitoring of the L-PBF process. The acoustic signals generated under different manufacturing parameters are recorded using a microphone during the L-PBF process. Acoustic features sensitive to manufacturing parameters are extracted from statistical features of the time-series, Fourier transform, power spectrum and ensemble empirical mode decomposition to probe more deeply into the acoustic signal. A subset of the most informative data features is selected and then used to train a support vector regression model to predict the manufacturing parameters with the average relative error of 4–9%. The prediction results with high accuracy demonstrate the effectiveness of the selected feature subsets and the feasibility of the process monitoring for L-PBF, which contributes to achieving intelligent monitoring and predictive maintenance.</p>