The purpose of this study was to evaluate the total success rates of mineral trioxide aggregate (MTA), ferric sulfate (FS), and formocresol (FC) as pulpotomy agents in primary molars. A randomized, split-mouth study design was used in 32 healthy 5- to 7-year-old children with 128 carious primary molars without clinical or radiographic evidence of pulp degeneration. The pulpotomy agents were assigned as follows: Group 1=MTA; Group 2=FS; Group 3=1:5 diluted Buckley's FC; and Group 4=zinc oxide eugenol (ZOE) base. Clinical and radiographic follow-up at 6, 12, and 24 months used the following criteria: pain; swelling; sinus tract; mobility; internal root resorption; and furcation and/or periapical bone destruction. The data were analyzed using chi-square. No significant differences in success rates were found among the groups at 6 and 12 months. Success rates in groups 1 to 4 at 24 months were 96%, 88%, 88%, and 68% respectively. There was a significant difference (P<.001) between the MTA and ZOE groups at 24 months. ZOE, as the only pulpotomy medicament, had a significantly lower success rate than MTA. No significant differences were observed, among the 3 experimental materials (MTA, FC, and FS) at 2 years follow-up.