Purpose: The purpose of this study was to evaluate the radiographic bone level and stability changes around early loaded chemically modified sandblasted and acid-etched implants with and without a machined collar. Materials and Methods: Seventy-two chemically modified sandblasted and acid-etched implants 4.1 mm in diameter and 8 mm in length were placed in six dogs. Thirty-six implants had no machined collar (NMC) and 36 had a 2.8-mm machined collar (MC). Resonance frequency measurements were obtained at placement and weekly for 3 weeks. All implants were loaded 21 days after surgery. Standardized periapical radiographs were obtained at baseline, at 3 weeks, and at 3, 6, 9, and 12 months. The radiographs were randomized and digitized, and linear measurements of the distance from the first bone-to-implant contact to the shoulder of the implant were performed at the mesial and distal aspects of each implant. For statistical analysis, mixed-model repeated-measures analysis of variance was used. Results: All implants achieved hard and soft tissue integration clinically and radio-graphically and were clinically immobile. From placement to week 3, the mean implant stability increased for MC implants by more than 5 ISQs and for NMC implants by more than 7 ISQs. Radio-graphically, there were significant differences between treatment groups beginning at 3 months. After 12 months of loading, the MC implants presented a mean bone loss of 1.00 mm and the NMC implants presented a mean bone gain of 0.11 mm. Conclusions: Chemically modified sandblasted and acid-etched implants without a machined collar presented bone gain, and implants with a machined collar showed bone loss after a 1 year following early (21-day) loading. The tendency toward a coronal apposition of bone observed under these conditions may be attributed to the osteoconductive properties of the chemically modified surfaces of these implants and to the absence of the machined collar. Schlagwörter: alveolar bone loss/etiology, dental implants, early loading, follow-up stability, osseointegration, resonance frequency analysis