The aim of this study was to examine and to compare alterations in the secretion of atrial natriuretic peptide (ANP) during different exercise-testing protocols in moderately trained men. Fifteen healthy male physical education students were studied (mean age 22·3 ± 2·5 years, training experience 12·3 ± 2·5 years, height 1·80 ± 0·06 m, weight 77·4 ± 8·2 kg). Participants performed an initial graded maximal exercise testing on a treadmill for the determination of VO(2max) (duration 7·45-9·3 min and VO(2max) 55·05 ± 3·13 ml kg(-1) min(-1) ) and were examined with active recovery (AR), passive recovery (PR) and continuous running (CR) in random order. Blood samples for plasma ANP concentration were taken at rest (baseline measurement), immediately after the end of exercise as well as after 30 min in passive recovery time (PRT). The plasma ANP concentration was determined by radioimmunoassay (RIA). The results showed that ANP plasma values increased significantly from the rest period to maximal values. In the short-term graded maximal exercise testing the ANP plasma values increased by 56·2% (44·8 ± 10·4 pg ml(-1) versus 102·3 ± 31·3 pg ml(-1) , P<0.001) and in the CR testing the ANP levels increased by 29·2% (44·8 ± 10·4 pg ml(-1) versus 63·3 ± 19·8 pg ml(-1) , P<0.001) compared to the baseline measurement. Moreover, the values of ANP decreased significantly (range 46·4-51·2%, P<0.001) in PRT after the end of the four different exercise modes. However, no significant difference was evident when ANP values at rest and after AR and PR were compared. It is concluded that the exercise testing protocol may affect the plasma ANP concentrations. Particularly, short-term maximal exercise significantly increases ANP values, while the intermittent exercise form of active and passive recovery decreases ANP concentrations.