SUMMARYThe proportion of sward area rejected for grazing under different stocking intensities in grass-white clover (Trifolium repens L.) swards continuously grazed by steers toconstant sward height was determined and the effect of these areas on clover stolon developmentwas investigated.In a randomized block field experiment, grazing treatments imposed from 1988 to 1991 comprised target sward surface heights (SSH) of 5, 7 or 9 cm from April to July followed by 7 or 9 cm from July to October (i.e. six treatments). The percentage area rejected was higher in 9 cm than in the shorter swards, the proportion of the total areas rejected increasing generally over the year. Within the rejected areas in spring the number of branches on the main axes of white clover plants was only 17% and the mean number of nodes per branch was 14% of those in grazed areas. Corresponding percentages in summer in rejected areas were 56 and 46% compared with grazed areas.In a glasshouse experiment involving stolons grown for 18–21 days, the difference between those from grazed and rejected areas in branch number per stolon and mean node number per branch was much reduced. Branching and rate of branch development (node number) were slower in stolons maintained under dark conditions.Mean growing point density over six sampling dates from summer 1990 to autumn 1991 was significantly higher in the 5 cm than in the 9 cm spring SSH treatment. Although rejected areashad only 34% of the growing point density of grazed areas in autumn, the latter comprised the majority of the total area in most sward height treatments and so reduced the overall effect of rejected areas on growing point density.