Introduction: Quality of chest compressions (CC) is key for patient’s sur-vival to cardiac arrest. Current resuscitation guidelines recommend CC rates between 100 and 120 per minute. Incorporating real-time feedback on CC rate in automated external defibrillators (AED) could enhance CC quality. In addition, automatic detection of pauses in CC (hands-off intervals) could allow continuous rhythm analysis during resuscitation. Aim: To design a method for computing CC rate and detecting CC pauses using the transthoracic impedance (TTI) signal acquired from the AED defib-rillation pads. Materials and methods: Thirty-five episodes with durations above 1000 seconds were randomly selected from a database of AED recordings collect-ed by Emergentziak-Osakidetza (Basque Country, Spain). Episodes con-tained the ECG and the TTI signals, which were resampled to 250 sps. Epi-sodes were randomly split into a training (10 episodes) and a test subset. A CC pause was annotated when CC were interrupted for more than 3 s. Our method computes the autocorrelation of the TTI signal using a 2-s analysis window. The position of the highest peak in the range 0.25-0.8 s is used to compute CC rate. A CC pause is detected when computed CC rate equals zero. The method was optimized using the training set. Results: With the test set, the method reported a sensitivity/positive pre-dictive value (PPV) in the detection of CC pauses of 97.64%/92.48%. Relia-ble CC rates were obtained with a sensitivity/PPV of 99.03%/98.61%. Unreli-able CC rates were reported only in 0.9% of the cases. CC rate measurement error was not computed due to lack of gold standard. Conclusions: A method based on the autocorrelation of the TTI signal al-lows reliable detection of CC pauses and estimation of CC rates, enhancing current AED functionality.