This study investigated the problem of the influence of 50 Hz magnetic field (MF) on human heart rate variability (HRV). The exposure system was a commercial device for magnetotherapy, generating field of the strength of 500 microT at the center of the coil, 150-200 microT at the position of human subjects\' heart and 20-30 microT at the position of subjects\' head. The exposure protocols, applied randomly, were either "half hour MF-off/half hour MF-on" or "half hour MF-off/half hour MF-off." The phonocardiographic (PhCG) signal of 15 volunteers were obtained during exposure and used for calculation of time-domain HRV parameters (mean time between heart beats (N-N), standard deviation of time between heart beats (SDNN), and the number of differences of successive beat-to-beat intervals greater than 50 ms, divided by the total number of beat-to-beat intervals (pNN50)) and nonlinear HRV measures (approximate entropy (ApEn), detrended fluctuation scaling exponents). The protocol MF-off/MF-on was applied in nine subjects. Repeated measures ANOVA (RMANOVA) performed for Mf-off/MF-off protocol indicated no statistical difference among four 15 min intervals of HRV data (P value >20% for all parameters except for N-N, where P = 3.7%). RMANOVA followed by the post hoc Tukey test performed for Mf-off/MF-on protocol indicated a statistically significant difference during MF on for N-N (8% increase, P <.1%), SDNN (40% increase, P = 1.1%), and pNN50 (110% increase, P <.1%). The results of the analysis indicate that the changes of these parameters could be associated with the influence of MF.