Several regulatory bodies have approved a health claim on the cholesterol-lowering effects of oat β-glucan at levels of 3·0 g/d. The present study aimed to test whether 1·5 g/d β-glucan provided as ready-to-eat oat flakes was as effective in lowering cholesterol as 3·0 g/d from oats porridge. A 6-week randomised controlled trial was conducted in eighty-seven mildly hypercholesterolaemic ( ≥ 5 mmol/l and < 7·5 mmol/l) men and women assigned to one of three diet arms (25 % energy (E%) protein; 45 E% carbohydrate; 30 E% fat, at energy requirements for weight maintenance): (1) minimal β-glucan (control); (2) low-dose oat β-glucan (1·5 g β-glucan; oats low - OL) or (3) higher dose oat β-glucan (3·0 g β-glucan; oats high - OH). Changes in total cholesterol and LDL-cholesterol (LDL-C) from baseline were assessed using a linear mixed model and repeated-measures ANOVA, adjusted for weight change. Total cholesterol reduced significantly in all groups ( - 7·8 (sd 13·8) %, - 7·2 (sd 12·4) % and - 5·5 (sd 9·3) % in the OH, OL and control groups), as did LDL-C ( - 8·4 (sd 18·5) %, - 8·5 (sd 18·5) % and - 5·5 (sd 12·4) % in the OH, OL and control groups), but between-group differences were not significant. In responders only (n 60), β-glucan groups had higher reductions in LDL-C ( - 18·3 (sd 11·1) % and - 18·1 (sd 9·2) % in the OH and OL groups) compared with controls ( - 11·7 (sd 7·9) %; P = 0·044). Intakes of oat β-glucan were as effective at doses of 1·5 g/d compared with 3 g/d when provided in different food formats that delivered similar amounts of soluble β-glucan.