Between February 1986 and November 1989, 166 patients who were candidates for a bone marrow transplantation entered a randomised controlled clinical trial to compare limited oral hygiene care (LIM) and intensive oral hygiene care (INT) in the prevention of mucositis. Randomisation was stratified on the initial oral status (good vs. bad IOS). Intensive oral hygiene care included an initial treatment of dental lesions and tooth and gum brushing during aplasia. Limited oral hygiene care excluded preventive dental treatment and gingival and tooth brushing. Mucositis was classified as absent, mild, moderate or severe, according to the clinical aspects of the different sites in the mouth and to two scales of pain evaluation. Of the 150 evaluable patients (75 in each group), 134 developed moderate/severe mucositis (64 in the INT group and 70 in the LIM group) (log-rank test P < 0.02). The superiority of intensive oral care was observed both in patients with and without total body irradiation (TBI) and in patients with a good or bad IOS; the observed risk of mucositis was reduced by 70% in each of these four subgroups. Duration of moderate/severe mucositis was, although not significantly, lower in the INT group (17 days, S.D. = 12) than in the LIM group (19 days, S.D. = 13). The median time of mucositis occurrence was 11 days in the INT group and 9 days in the LIM group. Contrary to a widespread belief, the percentage of documented septicaemia was not higher in patients who underwent intensive oral care. We conclude that, although statistically significant, the superiority of intensive oral hygiene care is not clinically impressive.(ABSTRACT TRUNCATED AT 250 WORDS)