Beagle dogs were given chloroform in a toothpaste base orally in gelatin capsules on 6 d/wk for 7 1/2 yr, followed by a 20-24 wk recovery period. Groups of 16 males and females received 0.5 ml/kg/d of the vehicle (toothpaste without chloroform) and 8 dogs of each sex remained untreated. Treated groups comprised 8 dogs of each sex remained untreated. Treated groups comprised 8 dogs of each sex, receiving doses equivalent to 15 and 30 mg CHCl3/kg/d in the toothpaste vehicle; another group of the same size received an alternative non-chloroform toothpaste (0.5 ml/kg/d). Eleven of the 96 dogs died during the study, only two of these being in the CHCl3-treated groups. The only significant toxic response during treatment was a moderate rise in serum enzyme levels (e.g. SGPT), reaching a peak in the sixth year of the study and probably corresponding to minimal liver damage. Few Palpable growths were noted while the dogs were alive. "Fatty cysts" were seen in the liver of several dogs at post mortem possibly associated with the chloroform treatment but the distribution of a nodular change in the liver was not obviously dose related. A small number of macroscopic and microscopic neoplasms were seen; one dog in each chloroform-treated group had a malignant tumour but there were no tumours in the liver or kidney of any dog. Overall, exposure to chloroform in a toothpaste base was not associated with any effect on the incidence of any kind of neoplasm. From this and related studies in mice and rats, it is concluded that repeated exposure to chloroform (3.5 percent) in toothpaste is unlikely to result in any hazard to human health.