Over the last decade, a variety of noninvasive techniques have been developed to monitor therapeutic ultrasound procedures in support of safety or efficacy assessments. One class of methods employs diagnostic ultrasound arrays to sense acoustic emissions, thereby providing a means to passively detect, localize, and quantify the strength of nonlinear sources, including cavitation. Real array element diffraction patterns may differ substantially from those presumed in existing beamforming algorithms. However, diffraction compensation has received limited treatment in passive and active imaging, and measured diffraction data have yet to be used for array response correction. The objectives of this paper were to identify differences between ideal and real element diffraction patterns, and to quantify the impact of diffraction correction on cavitation mapping beamformer performance. These objectives were addressed by performing calibration measurements on a diagnostic linear array, using the results to calculate diffraction correction terms, and applying the corrections to cavitation emission data collected from soft tissue phantom experiments. Measured diffraction patterns were found to differ significantly from those of ideal element forms, particularly at higher frequencies and shorter distances from the array. Diffraction compensation of array data resulted in cavitation energy estimates elevated by as much as a factor of 5, accompanied by the elimination of a substantial bias between two established beamforming algorithms. These results illustrate the importance of using measured array responses to validate analytical field models and to minimize observation biases in imaging applications where quantitative analyses are critical for assessment of therapeutic safety and efficacy.