Disclosed is a of monitoring the breathing of a user, comprising receiving a plurality of audio signals from a plurality of microphones and beamforming the plurality of audio signals into a beamformed audio signal. Observed breath-related parameters are derived from the beamformed audio signal and an output of or derived from the observed breath-related parameters is provided to a user of a client device. Breath metrics may be determined by comparing the observed breath-related parameters with reference breath-related parameters, and a breathing score may be derived from the breath metrics. The breathing score may be based on a combination of how closely an observed respiration of the user matches a prompted or desired respiration rate and an observed amount of time that the user's observed respiration rate is within a threshold value of the prompted or desired respiration rate.