Abstract. The effects of long-term sodium restriction on plasma atrial natriuretic factor (ANF) concentrations, and the role of baseline plasma ANF concentration as an indicator of changes in haemodynamics and left ventricular hypertrophy during this treatment were studied in 40 middle-aged previously untreated mildly to moderately hypertensive men and women in a 6-month controlled randomized study. The main emphasis of the treatment programme was to reduce daily sodium intake to less than 70 mmol. Mean sodium excretion decreased from 148 ± 74 mmol 24 h-1 to 79 ± 71 mmol 24 h-1 in the treatment group, but remained unchanged in the control group (173 ± 68 mmol 24 h-1 vs. 186 ± 62 mmol 24 h-1; P < 0.01 for the difference in changes between the groups). Mean plasma ANF concentrations in the treatment group were 52.4 ± 20.7 (median 50) pg ml-1 at baseline and 38.7 ± 26.3 (median 42) pg ml-1 at 6 months, and the corresponding values in the control group were 55.5 ± 20.5 (median 50) pg ml-1 and 46.1 ± 32.4 (median 50) pg ml-1, respectively (P = NS for the difference in changes). The ANF concentration decreased from 70 ± 14 pg ml-1 to 32 ± 26 pg ml-1 in treated subjects with a high baseline plasma ANF concentration (> 50 pg ml-1), but increased from 37 ± 11 pg ml-1 to 45 ± 27 pg ml-1 in subjects with a low baseline plasma ANF concentration (< 50 pg ml-1) (difference in changes P < 0.001). Compared with treated subjects with low baseline plasma ANF levels and with controls, treated subjects with high baseline plasma ANF levels showed a decrease (P < 0.05) in interventricular septal and left posterior wall thicknesses, in relative wall thickness, and in peripheral resistance. These results suggest that in mildly to moderately hypertensive subjects long-term sodium restriction decreases high plasma ANF concentrations concomitantly with regression of concentric left ventricular hypertrophy, probably as a result of changes in haemodynamics.