Aim. Interoception reflects the perception, integration and modulation of visceral signals. Currently, there is no established universal approach to cardiac interoception assessment allowing its inclusion in the diagnostic protocol for various diseases. The aim was to compare three tests of interoception self-assessment based on heart palpitations within one day in healthy subjects.Material and methods. Eighty healthy volunteers took part in the study. To assess cardiac interoception, a test with button pushes at the moment of palpitations, a feedback test in the form of synchronous and asynchronous sound signals, a heartbeat counting task at specified time intervals were conducted with simultaneous electrocardiographic recording. The obtained data were compared with each other (Wilcoxon test). Its pairwise dependence (Spearman coefficient) and dependence on body parameters (sex, age, body mass index, fat mass, anxiety) were assessed (univariate linear regression).Results. The accuracy of the heartbeat counting task turned out to be significantly higher than in the button- pressing test (p=0,005). However, both estimates correlate with each other (r=0,66, p=0,003). A positive dependence of test accuracy on sex was found for the heartbeat counting task (p=0,021) and a negative dependence on fat mass for button- pressing task (p=0,032).Conclusion. The study found that volunteers' estimates of heartbeat accuracy varied across tests and were also associated with sex and body composition. The heart rate counting task and the button- pressing task are the most suitable for interoception assessment.