This paper provides a summary of routine evoked potential tests used with rats, with elaboration on the cochlear microphonic portion of the auditory brainstem response, the effects of chemicals on high frequency (above 40 Hz) components of the somatosensory evoked potential, on cerebellar recording of sensory evoked potentials, and on central conduction time. An alternative to peak-valley amplitude and latency measurements is discussed, wherein a computer analyzes evoked potentials for differences from control in waveform shape, latency, and power. Since multiple use of statistics is common, resulting in an inflated false positive rate, an alpha criterion of less than 0.05 is recommended. Instead of dividing alpha by the number of statistical tests (Bonferroni), a less severe correction of dividing alpha by the square root of the number of tests is proposed.