Cardiac monitors determine heart rate by counting the number of beats in a given time interval. The normal heart, however, does not beat at a constant rate. Instead, there is a continuous change in heart rate on a beat-by-beat basis. This is termed the instantaneous heart rate and it represents the projected rate per minute that the heart would beat if only one R-R interval (the time between sequential R waves) was repeated throughout a 60-second period. Calculation of the instantaneous heart rate for each heart beat (R-R interval) produces a pattern that demonstrates the variability in heart rate. This instantaneous heart rate pattern was prospectively studied in 102 patients admitted to a neurosurgical intensive care unit. Short-term (STV) and long-term (LTV) heart rate variability were compared to the Glasgow coma scale as a method for patient assessment. LTV seems to be the most useful heart rate parameter in the clinical setting, and both STV and LTV performed better in the serial evaluation of patients. Two postulations found in the heart rate literature were not borne out in this study. First, we did not find a strong correlation between elevated intracranial pressure and decreases in heart rate or variability, as previously reported by Lowensohn et al. Second, the morphological classification of heart rate patterns described by Evans in his study of head-injured patients did not carry the same prognostic value when applied to this broad spectrum of patients with a variety of acute neurological disorders. Heart rate and its variability is a simple parameter to monitor. In comparison to multimodality evoked potential testing, heart rate monitoring is quicker, less cumbersome, and requires less technical skill. It can provide useful clinical information, especially for frequent serial examination of brain stem-injured patients. Periodicity in the heart rate pattern was noted in 33% of the patients studied. This technique may provide valuable information about postulated brain stem oscillatory centers governing heart rate, and it should be useful in the future investigation of neurocardiac interrelationships. (Neurosurgery 19:356-362, 1986)