Introduction: ECG is a biological signal specific for each person that is hard to create artificially. Therefore, its usage in biometry is highly investigated. Since now, several articles about ECG biometry have been published and the results seem to be promising. It may be assumed that in the future, ECG for biometric purposes will be measured by wearable devices. Therefore, the quality of the acquired data will be worse compared to ambulatory ECG. In this study, we proposed and tested three different ECG-based authentication methods. Materials and Methods: The data were measured by Maxim Integrated wristband. In total, 29 participants were involved. The created database contains 810 recordings each with duration of 20 s and sampling frequency of 512 Hz. Three participants have over 90 recordings each. Consequently, three different approaches were designed for the person authentication. The first method extracted 22 time domain features - intervals and amplitudes from each heartbeat and Hjorth descriptors from an average heartbeat. The second method used 320 features extracted from the wavelet domain. For both methods a random forest was used as a classifier. The deep learning method was selected as the third method. Specifically, the 1D convolutional neural network with embedded feed-forward neural network was used to classify the raw signal of every heartbeat. Results: Each method was tested in a scenario, when one person with a high number of measurements was tested against all other individuals. The first method reached an average false acceptance rate (FAR) 7.11% and false rejection rate (FRR) 6.49%. The second method reached FAR 7.34% and FRR 20.04%. The third method reached FAR 0.57% and FRR 0.00%. Conclusion: The achieved results confirm a great potential for biometry purposes. Especially the results of the deep-learning method are very promising.