This is a dataset of single-trial ABR data with ABR thresholds judged by expert human raters. It was used to test the ABRpresto algorithm, which is reported here and in submission. It contains a total of 7,857 ABR waveform stacks from 351 mice. It includes mice with normal hearing and hearing loss from multiple strains and genotypes, totaling 123,140 unique ABR waveforms. The data is in psi format, example code in the ABRpresto github illustrates how to convert it to csv if desired. It also includes two csv files: "ABRpresto thresholds 10-29-24.csv" contains the resulting thresholds generated by running the dataset through the ABRpresto algorithm. "Manual Thresholds.csv" contains the thresholds picked by expert human raters using an interactive GUI similar to that available here. It also contains min_level and max_level columns that describe the minimum and maximum levels tested for each dataset. See Example_plot_performance.py for an example of how to use these files to plot algorithm performance. Data were collected using a customized version of Psiexperiment, a plugin-based framework for auditory experiments (Buran and David, 2020, https://zenodo.org/records/3661386). Five-millisecond tone-pips with a 0.5 ms rise-fall time delivered at 81/s in alternating polarity were used for frequency-specific measurements of hearing function. Sound levels were tested in 5 dB steps, at either a 15–105 dB, 15–85dB, or 65–105 dB sound pressure level (SPL) range. Stimulus frequencies ranged from 4 to 45.3 kHz in half-octave steps. Stimuli were presented in interleaved order such that a train of tone-pips containing a single presentation of each level and frequency was repeated in an interleaved ramp paradigm (as described in Buran et al. 2020). Electrical signals were recorded through a digitizing amplifier (TDT RA4PA or Medusa4Z). Responses from 512 repetitions were recorded, and individual response to each trial and averaged responses were stored for further analysis. All experiments were approved by the Animal Care and Use Committee of Decibel Therapeutics.