We describe a quantitative model capable of explaining the results of all reported investigations of the counting of interaural and monaural click sequencies. The model is developed by means of three convergent operations: (a) reanalyzing absolute-estimation data of apparent repetition rates of interaural and monaural sequences, (b) deriving interaural and monaural counting times from numerosity-judgement data, and (c) analyzing the time that observers needed to respond to the end of interaural and monaural sequences. The combined evidence demonstrates that the perceived onset asynchrony (POA) between interaural events is 24 msec longer than that between monaural events. The model has three components: (a) a "stimulus clock," which represents the stimulus onset synchrony (SOA) between events; (b) a "memory clock," which represents the POA between events, and (c) a "counting clock," which represents the counter increment time. The transfer functions between the three clocks are deduced from empirical data. Other proposals to explain interaural click counting results (attention switching, streaming by locus, counterincrement deficit) are discussed and rejected.