OBJECTIVE To compare the clinical and radiological outcomes following two different, single visit vital pulp therapy techniques, in cariously exposed primary molar teeth. METHODS A paediatric dental clinic within the Dental Hospital, Newcastle upon Tyne, UK. METHODS Fifty two child patients were sequentially enrolled in the clinical investigation, 26 males and 26 females with an age range of 3.3-12.5 years. Primary molar teeth requiring vital pulp therapy were randomly allocated to either the formocresol group (F) or the calcium hydroxide group (C). The total number of teeth treated was 84. METHODS Recruitment was on the basis of strict inclusion criteria. Coronal pulp amputation was prescribed only in teeth with vital, cariously exposed pulp tissue. Treatment was undertaken between October 1994 and December 1996. All cases were reviewed using predefined clinical and radiological criteria. The statistical tests used were logistic regression of a triple nested data structure, chi-squared analysis of equality of treatment and probability of success with relation to subject age. RESULTS Eighty-four cariously exposed primary molars required vital pulp therapy. Forty six (55%) teeth were included in the F group and 38 (45%) allocated to the C group. Five teeth were lost to follow-up, leaving 79 teeth: forty four (56%) in group F and 35 (44%) in group C. Eighty four percent (37/44) of teeth treated with formocresol and 77 percent (27/35) treated with calcium hydroxide were classed as clinically and radiographically successful at the cut-off date, December 1997, after a mean clinical review of 22.5 months (range 6.1-38.5 months) and a mean radiographic review of 18.9 months (range 1.3-36.9 months). CONCLUSIONS This investigation confirms the clinical efficacy of a one-fifth dilution of Buckley's Formocresol as an agent in pulp treatment of cariously exposed, vital primary molar teeth. However, calcium hydroxide in its pure, powder form is a clinically acceptable alternative when combined with strict selection criteria for this method of restorative care. There was a statistically insignificant difference in successful clinical and radiological outcome between the two treatment groups. Success was unrelated to the duration of time taken to achieve haemostasis and the presence or absence of bleeding after placement of the medicament.