In the period 1963-1974, 82 monks and 48 nuns from five Dutch and Belgian Trappist monasteries each participated in two or more out of nine different trials designed to test the effect of 58 different fat-modified diets on serum cholesterol. We analysed these data to quantify the extent to which healthy, normolipaemic subjects differ in the responsiveness of their serum cholesterol to a change in dietary fatty acid composition. Statistically significant between-person variance components (SD2p) were found in the serum cholesterol responses for the whole group (SD2p = [0.20 mmol l-1]2), for the men (SD2p = [0.24 mmol l-1]2) and for those women who participated in three or more trials (SD2p = [0.14 mmol l-1]2). The between-person variation (expressed as SD) was on average only half as large as the within-person variation in response when the same subject was challenged repeatedly. It is concluded that medically significant differences in responsiveness to fat-modified diets exist in both men and women. However, few subjects fail entirely to respond to a change in dietary fatty acid composition. In addition, the large within-subject variability makes it difficult to identify hypo- and hyperresponders.