Alternating use of the original method of internal drill cooling with manual external cooling, one IMZ implant was placed in the diaphysis (compact bone) and metaphysis (spongy bone) of the left distal tibia in each of six sheep. After polyfluorochrome sequential labeling, two animals survived for 4, 8, and 16 weeks without functional loading of the implants. The drill holes and bone-implant reactions were evaluated in undecalcified ground sections and microradiographs, using computer-aided histomorphometry. At least 30% of all drill holes were incongruent, but up to 25% of the implant surfaces had primary bone contacts. Four weeks after implantation, distinct differences in the bone reactions could be determined between the experimental conditions. After both cooling methods, osteoclastic resorption had extended up to 0.5 mm into peri-implant compact bone and up to 0.18 mm into spongy bone. Based on the percentage of newly formed bone-to-implant contacts, external cooling proved superior at all superficial drill hole levels in compact bone and in all spongy bone beds, while internal cooling was only better at the deeper drill hole levels in compact bone. After 8 and 16 weeks, peri-implant bone remodeling led to further increased new bone-implant contacts, irrespective of incongruent drill holes or cooling methods. At all time periods, abraded titanium particles were observed within interfacial foreign body cells. Conclusions are: (1) additional external cooling seems beneficial for any internal cooling, particularly in compact bone; (2) spongy bone apparently tolerates drilling heat better and reacts with more and earlier new bone formation; (3) the IMZ drill set cannot guarantee a congruent drill hole; (4) the titanium plasma-sprayed coating is endangered by mechanical abrasion. Schlagwörter: bone structure, drill cooling, endosseous implant, histomorphometry