BACKGROUND: Studies revealed long-term associations between noise exposure and cardiovascular health, but the underlying short-term mechanisms remain uncertain. OBJECTIVES: To explore the concomitant and lagged short-term associations between personal exposure to noise and heart rate variability (HRV) in a real life setting in the Ile-de-France region. METHODS: The RECORD MultiSensor Study collected between July 2014 and June 2015 noise and heart rate data for 75 participants, aged 34-74 years, in their living environments for 7 days using a personal dosimeter and electrocardiography (ECG) sensor on the chest. HRV parameters and noise levels were calculated for 5-min windows. Short-term relationships between noise level and log-transformed HRV parameters were assessed using mixed effects models with a random intercept for participants and a temporal autocorrelation structure, adjusted for heart rate, physical activity (accelerometry), and short-term trends. RESULTS: An increase by one dB(A) of A-weighted equivalent sound pressure level (Leq) was associated with a 0.97% concomitant increase of the Standard deviation of normal to normal intervals (SDNN) (95% CI: 0.92, 1.02), of 2.08% of the Low frequency band power (LF) (95% CI: 1.97, 2.18), of 1.30% of the High frequency band power (HF) (95% CI: 1.17, 1.43), and of 1.16% of the LF/HF ratio (95% CI: 1.10, 1.23). The analysis of lagged exposures to noise adjusted for the concomitant exposure illustrates the dynamic of recovery of the autonomic nervous system. Non-linear associations were documented with all HRV parameters with the exception of HF. Piecewise regression revealed that the association was almost 6 times stronger below than above 65 Leq dB(A) for the SDNN and LF/HF ratio. CONCLUSION: Personal noise exposure was found to be related to a concomitant increase of the overall HRV, with evidence of imbalance of the autonomic nervous system towards sympathetic activity, a pathway to increased cardiovascular morbidity and mortality.