Using 22 erupting human third mandibular molars, this study investigated: (1) the relationship between the morphology of the interlobal groove and the histological features of caries, (2) the ultrastructural features of the interlobal groove contents, and (3) whether the morphology of interlobal grooves influences the viability of the microorganisms. Posteruption time was classified as (1) <3 months (7 teeth), (2) 3-12 months (8 teeth) and (3) >12 months but <18 months (7 teeth). The teeth were pre-and postfixed, and buccolingual sections were prepared. The mesial/distal interlobal groove on 116 sections was classified as fissure-like or groove-like. All sections were embedded and ground, and microradiographs were made. The interlobal groove contents were analyzed on two sections from each tooth, one with a fissure-like and one with a groove-like morphology, giving a total of 44 sections. Eighteen teeth (82%) had evidence of caries. Severity of caries was associated with length of posteruption time. 'Fissures' were not more prone to caries than 'grooves'. Tn about 70% of the 'fissures' caries had penetrated deepest at the entrance rather than in the deeper parts. LM and TEM examinations from 1 tooth disclosed erythrocytes, enamel organ cells and pioneers of microorganisms in the interlobal grooves. All other interlobal grooves were occupied by microorganisms, mainly gram-positive coccoids and small rods. Viable microorganisms were seen at the entrance as well as at the bottom part of the 'grooves'. In 'fissures' viable microorganisms were primarily observed at the entrance, while in deeper parts the microorganisms were less viable or dead. A moderate association was observed between the viability of the microorganisms and differences in the depths of caries penetration in the fissures. Thus, the internal morphology of the interlobal grooves influenced the conditions for bacterial growth and this determined the location for caries progression within the groove-fossa system.