The Class III Posterior Composite Restoration

Dentistry Today


The class III restoration is nominally a “one-surface” restoration on the proximal contacting surface of the tooth. It is generally formed with composite resin or ionomer cement. Class III restorations are consistent with the philosophy of “minimal invasion,” as the only healthy tooth structure that is removed is that necessary for access to caries and retention of the restorative material. Retention is generally provided by small internal undercuts in the dentin and/or bonding of composite resin to etched bevels of the peripheral enamel.

Prior to the development of composite resin and acid-etch techniques, proximal tooth-colored restorations were made with silicate cement, and they were restricted to anterior lesions mesial to the midline of the cuspid. Proximal lesions on the distal of the cuspids and the posterior teeth were generally treated with class II metal restorations. This was thought necessary in order to prevent excessive mesial drift of the posterior teeth that would likely occur as silicate cement slowly dissolved over time.

Composite resins, in comparison, are not soluble and do not exhibit similar in vivo degradation. They are widely used as a class II posterior restorative material due to their clinical durability. Although interproximal wear of composite resin has been noted in at least one clinical study,1 the author is unaware of any reports in the dental literature of mesial drift of posterior teeth when they are restored proximally with contemporary composite resins. The absence of such reports coupled with low rate of occlusal wear exhibited by contemporary composite resins suggests that there is no reason to limit class III composite restorations only to anterior teeth.

Preparation designs recommended for class II composite restorations have undergone an evolution from Black’s class II amalgam cavity preparation to less invasive proximal “slot” preparations that eliminate the isthmus and dovetail, thought by many as necessary for the retention of nonadhesive amalgam.2-6 Since slot preparations access caries through the occlusal marginal ridge of the tooth, the challenges of restoring the proximal contact and marginal ridges remain.

These difficulties can be reduced or overcome by restoring posterior proximal carious lesions with composite resin using preparations that access the lesion through the buccal or lingual enamel rather than through the occlusal surface. The buccal approach is generally more accessible, and the smooth buccal surface of the tooth is much easier to restore anatomically than the occlusal surface. Since “virgin” proximal lesions generally occur slightly apical to the contact area, it is often not necessary to completely break the contact in order to successfully restore the tooth. This significantly reduces the challenges of restoring a tight contact. In these cases (as with anterior teeth) the composite resin can often be contained using only a transparent Mylar strip rather than specialized matrix bands.


Figure 1. Preoperative radiograph of proximal lesion. Figure 2. Preoperative buccal view.
Figure 3. Preoperative occlusal view. Figure 4. Buccal view of isolated tooth.
Figure 5. Initial access with a long, narrow diamond bur. Figure 6. Initial access to the lesion.
Figure 7. Visually complete caries removal. Figure 8. Caries dye reveals remaining caries.
Figure 9. All caries is removed. Figure 10. Applying ionomer cement to cover all dentin.

The patient presented with an interproximal carious lesion on the distal of the lower premolar (Figure 1). Radiographically, it penetrated through the enamel into the dentin. The lesion was barely clinically detectable with an explorer and was invisible to the naked eye (Figures 2 and 3).

The teeth were anesthetized and isolated with a rubber dam with a sufficient number of holes for convenient access. An interproximal wedge was placed to protect the interdental dam and gingival papilla and  provide slight separation of the teeth (Figure 4). Access to the carious lesion was gained through the buccal enamel using a long, narrow tapered high-speed diamond bur (Figure 5). Care was taken to avoid contacting the adjacent tooth with the bur (Figure 6). Although it was not done in this case, placing a piece of metal matrix band material between the teeth is a useful way to protect against accidental abrasion of the adjacent tooth. The access opening was enlarged sufficiently to permit convenient assessment and removal of all caries (Figure 7). Caries detecting dye was used to ensure complete removal of decay (Figures 8 and 9).

The restoration may be placed by either the “total-etch” technique using dentin-bonding agents prior to filling the lesion with composite resin, or by a “lamination” technique where a cement base is placed over all dentin, followed by bonding an external layer of composite resin to the etched, beveled peripheral enamel (Figure 10). The latter method is preferred by the author due to its simplicity and history of postoperative comfort. Any cement may be used as a dentin-replacement base, but fast-setting glass ionomers (eg, Fuji IX [G-C America], Ketac Molar [3M ESPE]) or light-cured resin ionomers (Fuji II l.c. [G-C], Vitrebond [3M ESPE]) are preferred. These cements adhere to tooth structure without the need for acid etching. The cement should cover all exposed dentin and terminate at the DEJ of the prepared cavity. In this situation, there is no need for high bond strength of the cement to dentin because of the strong bond of composite to the peripheral enamel.

If light-activated resin-ionomer cements are used, the dentist should be aware that some products produce substantial exothermic heat when light cured for more than 10 seconds. These cements should be applied in increments, limiting initial curing to 10 seconds or less. Self-curing cements do not liberate heat and may be placed in bulk.

Once the cement has initially hardened, any excess or overflow at the perimeter of the preparation is removed from the enamel, and the peripheral enamel is beveled about 1 mm using a fine-grit needle-shaped diamond (Figure 11). In some cases it may be desirable not to bevel the enamel that remains at the contact in order to avoid completely opening the contact.

Figure 11. The enamel is beveled. Figure 12. Etching the beveled enamel.
Figure 13. Applying unfilled resin. Figure 14. Remaining preparation is filled with composite; light cured 20 seconds on both the buccal and lingual.
Figure 15. Finished restoration. Figure 16. Finished restoration.
Figure 17. Postoperative radiograph of the restored tooth.

After beveling, a clear sectional matrix band or strip was placed interproximally and secured with a wedge. The enamel was etched for 5 to 20 seconds with 37% phosphoric acid (Figure 12). The acid was then rinsed and the tooth was thoroughly dried to reveal the characteristic frosty appearance of etched enamel. An unfilled resin (eg, D/E resin [Bisco]) was applied and blown to a thin film with compressed air (Figure 13). It was light cured for 20 seconds to establish a strong bond with the enamel. (Note: Dentin-bonding primers may be used in lieu of the unfilled enamel-bonding resin, but they are not recommended due to their greater cost and the presence of solvents that interfere with composite polymerization if they are not completely evaporated.)

While manually holding the wedged matrix band against the lingual surface of the tooth, the remaining cavity preparation was filled with a single increment of composite resin. The matrix band was pulled buccally around the composite adapting it to the beveled enamel. It was light cured for 20 seconds on both the buccal and the lingual before removing the band (Figure 14).

After light curing, the band and wedge were removed, permitting finishing of the margins. The proximal margins were finished with a series of rotary discs. The gingival margin was first trimmed with a No. 12 scalpel, followed by smoothing with an abrasive finishing strip (Figure 15). When it was complete, the restoration was visually undetectable (Figure 16), the proximal contact was solid, and there had been no alteration of the occlusal surface of the tooth. A postoperative radiograph confirmed the restoration (Figure 17).


This conservative method of restoring proximal carious lesions in posterior teeth is not offered as an “alternative” to more familiar methods. Rather, in the author’s view, anyone who embraces the philosophy of “minimal invasion” and respects the preservation of sound tooth structure should first consider a class III composite resin restoration for aesthetic treatment of proximal caries in posterior teeth before using any other more aggressive or complex cavity design. In the practice of conservative dentistry, the author feels that the conventional class II approach is the less desirable alternative.


1. Wendt SL Jr, Ziemiecki TL, Leinfelder KF. Proximal wear rates by tooth position of resin composite restorations. J Dent. 1996;24:33-39.

2. Leinfelder K. A conservative approach to placing posterior composite resin restorations. J Am Dent Assoc. 1996;127:743-748.

3. Spreafico R. Direct and semi-direct posterior composite restorations. Pract Periodontics Aesthet Dent. 1996;8:703-712.

4. Miller MB. Rest in peace G.V. Black: part 1. Pract Periodontics Aesthet Dent. 1997;9:757-758.

5. Albers H. Tooth-Colored Restoratives. 8th ed. Santa Rosa, Calif: Alto Books; 1996: (chapter 11) 1-12.

6. Ritter AV. Posterior resin-based composite restorations: clinical recommendations for optimal success. J Esthet Restor Dent. 2001;13:88-99.

Dr. Lacy is an associate professor in the Department of Preventive and Restorative Dental Sciences at the University of California School of Dentistry, San Francisco, Calif. He can be contacted at (415) 476-5431 or