Background: Parkinson's disease is a common neurodegenerative disease, and while early detection of cardiovascular symptoms may aid in early diagnosis, there are no definitive markers to assess cardiovascular autonomic damage in Parkinson's disease. Methods: Twenty Sprague Dawley rats were randomly assigned to the experimental (n=12) and control (n=8) groups, and a model of Parkinson's disease was created in the experimental group by stereotaxic injection of rotenone into the substantia nigra compacta and ventral tegmental area. Telemetry implantation was carried out after successful modeling. Equal dosages of saline were injected into control animals at the same places. In both groups of rats, ECG, blood pressure, core temperature, and activity were recorded using a Data Science International implantable physiological signal telemetry device. The time domain, frequency domain, and non-linear analysis were used to examine the blood pressure and heart rate variability of the two groups of rats. Results: The experimental group had higher Detrended Fluctuation Analysis (DFA) of blood pressure signals (systolic, diastolic, and mean arterial pressure), mean arterial pressure, normalized high frequency power (nHF) than the control rats (p<0.05). The experimental group had lower sample entropy (SampEn) of blood pressure signals (diastolic and mean arterial pressure), root mean squared successive differences (RMSSD), normalized lower frequency power (nLF) and total power than the control group (p<0.05). However, the standard deviation, coefficient of variation, and continuous variation in the linear analysis of continuous blood pressure signal between the two groups were not statistically significant (p>0.05). Conclusion: The rats in the rotenone model had significant autonomic dysfunction, and non-linear analysis approaches such as detrended fluctuation analysis and sample entropy were able to discern the diseased condition of the rats more sensitively while processing the continuous blood pressure data.