An evaluation of noise metrics for predicting human perception of sonic booms was performed. Twenty-five metrics were chosen from standards and from the literature in an effort to include all potentially relevant metrics. Three different datasets of sonic boom waveforms and associated human response were chosen to span a variety of signals, including traditional N-waves with various shock shapes and rise times, and predicted waveforms from designs of low-boom aircraft for a variety of aircraft sizes. These datasets were derived from laboratory studies conducted in sonic boom simulators at NASA Langley Research Center and JAXA. Simulations of booms experienced in both outdoor and indoor environments were included by using different facilities at NASA or modifications to facility configurations at JAXA. American and Japanese test subjects participated at NASA and JAXA, respectively. Ratings of loudness using a magnitude estimation technique and ratings of annoyance using a category line scaling method are included. The evaluation consists of linear correlations of human response data with the objective noise metrics. Results are presented for each study, and eight metrics are suggested for further analysis.