The heart rate variability (HRV) signal carries important information about the systems controlling heat rate and blood pressure, mainly elicited by autonomic nervous system (sympathetic and parasympathetic) controls. The present paper illustrates methods of HRV signal processing by using autoregressive (AR) modeling and power spectral density estimate. The information enhanced in this way seems to be particularly sensitive in discriminating various cardiovascular pathologies (hypertension, myocardial infarction, diabetic neuropathy, etc.). This method provides a simple non-invasive analysis, based on the processing of spontaneous oscillations in heart rate. Particular emphasis is directed to the algorithms used and to their direct application by using proper computerized techniques: only a few paradigmatical examples will be illustrated as preliminary results.