The cariogenicity of Ferrero Milchschnitte and apple was tested in rats using a frequency-controlled feeding machine. Twelve litters of Osborne-Mendel rats orally infected with Streptococcus sobrinus OMZ-176 and Actinomyces viscosus Ny-1 were used. Four treatment groups were tested: (1) cooked starch/sucrose 1:1 as positive control; (2) Ferrero Milchschnitte; (3) apple; (4) gastric intubation only as negative control. Groups 1-3 were additionally fed by gastric intubation with essential nutrients. Rats on Milchschnitte and apple had less fissure and smooth-surface caries than rats on a starch/sucrose diet, but significantly more than the negative controls. The Milchschnitte group exhibited less plaque extent than the positive and the negative control. The apple group showed a higher plaque extent and massively extended erosions on molar surfaces. The CFU means of S. sobrinus were decreasing from treatment 1-4. Although differences between treatments were not significant, they were associated by concomitantly decreasing caries incidence. The test foods were also assessed using intraoral plaque-pH telemetry in 6 human volunteers. Both Milchschnitte and apple were acidogenic in human telemetry. They can therefore not be regarded as safe for teeth.