The aim of the present study was to investigate the effects of training variations on 24-hr urinary noradrenaline (NA) and adrenaline (Ad) levels and the adrenaline/noradrenaline (Ad/NA) ratio to search for a possible relationship between catecholamine excretion, training, and performance in highly trained swimmers. Fourteen swimmers (5 female and 9 male) were tested after 4 weeks of intense training (IT), 3 weeks of reduced training (RT), and 5 weeks of low training (LT). At the end of each period, the swimmers performed their best event at an official competition. Individual performances were expressed as percentage of the previous season's best performance. The changes in NA levels after 4 weeks of IT were negatively related to changes in training volume (r=-0.70, p<0.01) and total training load (r=-0.68, p<0.02). NA levels measured at the end of IT were positively related to changes in performance after three weeks of RT (r=0.77, p<0.01). The percentage changes in performance between RT and LT were related to NA levels at the end of RT (r=0.60; p<0.04). Ad/NA ratios and Ad were related to performance (r=0.58, p<0.01; r=52, p<0.01; respectively). The differences in Ad/NA ratios and Ad between two consecutive competitions were related to the differences in performance (r=0.59, p<0.01; r=0.49, p<0.01; respectively). 24-hr NA and the Ad/NA excretion ratio were related to both training variations and performance. Thus, 24-hr NA levels and Ad/NA ratio may provide useful markers for monitoring training stress in elite swimmers.