Influences of age, sex and audiogram on ABR latencies have been studied. Using a mathematical model of audiogram, the analysis of data finds a relative importance of age, sex and audiogram similar to previous studies. Audiogram slope was correlated only with the latencies of waves I and V, and multiple regression analyses indicate the slope effect is relatively weak compared to that of the other parameters. Using a mathematical model of audiogram did not improve the ABR variance across subjects, underscoring the need to discover other relevant variables to explain ABR latency.