Introduction: The ECG measurements performed outside the clinical settings are highly susceptible to noise which may affect their clinical interpretation. Among other noise types, the broadband electromyographic (EMG) noise spectrally coincides with the QRS complex, which makes its removal particularly challenging. A number of EMG-noise removal techniques have been developed. However, evaluation and comparison of denoising methods remains problematic due to the nonexistence of a set of genuine noise-free signals. Here, we present a SimEMG database composed of ECG signals recorded with and without EMG noise by a novel acquisition method. Method: The SimEMG acquisition relies on the assumption that the potential at every point along the arm is constant when muscles are relaxed. If the EMG noise is generated locally in hands, a potential difference between electrodes placed on two hands will include the EMG noise, while a potential difference between electrodes placed on the shoulders (deltoid muscle), distant from the noise source, will not. Using SimEMG acquisition method, we obtained the SimEMG database, now available as open source. Results: The database contains 147 signals in total, 37 out of which are noise-free and 110 noise-contaminated single-lead recordings. The recordings are 30 seconds long and were obtained by a sampling rate of 500 Hz. They are recorded on 14 healthy subjects, 5 males and 9 females, aged 40±13. The average signal-to-noise ratio (SNR) of the noise-contaminated signals, was 8.53±5.5 dB. Most of them had SNR < 8 dB, while only 7 recordings had SNR > 16 dB, indicating a high overall noise level. Conclusion: The SimEMG database could be used as a reference database for the evaluation of the EMG noise removal methods.