The definition of this treatment category is  a restoration accessing the lesion without completely cutting away the marginal ridge. The finish line is on the marginal ridge. Occlusion must not be on the marginal ridge for this preparation to be indicated. The occlusal surface, if caries-free, is not included,or, if lightly carious,  Fissure Sealant or Fissurotomy procedures can be performed, isolating the CDG from the proximal carious restoration.

This conservative approach to proximal  structure, paired with the above conservative occlusal treatment  renders a treated tooth with greater transverse (buccal-lingual) integrity. The significant benefit relative to cutting an isthmus and joining the proximal to the occlusal lesion, as in a standard GVB preparation, is a lessened likelihood of lingual cusp fracture over the long term, 25 to 40 years. See Biomimetic Dentistry

The core information for the box preparation is found at   Class II Outside-Inwards box Preparation.

What follows is the adaptations to the basic outside-In preparation for the incipient proximal lesion.


The slot restoration is not an easy restoration. Its finesse requires operator skill. Perfect control of the handpiece is needed, and the margin for error is narrow.

  • Penetrate tooth conservatively from the occlusal..Make initial penetration with a new 330 bur,  opening the marginal ridge only to the crest of its curvature. Then use a new 169L to widen the cut buccally and lingually towards the point necessary to break proximal contact, at the depth necessary to form a box. Essentially at this point the marginal ridge has been undermined but no definitive box form has been made and the proximal walls have not been defined. The axial wall is just barely inside the DEJ. A shell of enamel remains intact surrounding the lesion. The adjacent tooth has been completely spared from iatrogenic damage.
  • Follow DEJ inside tooth to correct gingival height: Correlate the necessary gingival descent of the prep by holding the bur against  a radiograph to establish the required depth of the gingival margin. This should be below the contact point with the adjacent tooth and below the extent of caries on the radiograph.
  • Why not use a 1169L bur for these cuts ? Does not the round end produce rounded line angles, which are more desirable for reducing stress concentration and also for adapting composite in placement? The answer: The tactile qualities of the 1169 bur are not nearly as good as the 169L, and tactile feedback is critical to keeping this preparation under control.
  • Outside-in cutting of box walls: Bring the 169L bur outside the tooth to the buccal embrasure, and visualize what path the bur tip must follow in order to meet the occlusal cut made within the tooth. Ensure that the bur is still razor sharp. Lay the patient’s head over to the side so you are working with direct vision. If possible, use both hands to triangulate the handpiece with multiple finger rests.
  • Initial “Outside-in” cut:Make the initial cut only one third of the gingival-occlusal box height. There will be a “peninsula” of enamel remaining between the outside of the tooth and the  previously hollowed marginal ridge. As that is transited, the bur will “fall into” the void behind the enamel shell.
  • It  helps to lean the handpiece off-vertical, with the handpiece head inclined toward the center of the tooth. This restrains the otherwise pernicious tendency of the bur to excessively encroach upon the proximal arm of the cusp.
  • Second “outside-in” cut: Drop the bur by  2 mm of depth towards the gingival, and repeat this “outside-in” cut. Again, slowly, low rpm, light touch, triangulated finger rests.
  • Third “Outside-in” cut”Repeat with a third cut to the desired gingival level if needed.
  • Always monitor the cusp arm, because the taper of the bur wants to encroach on the cusp arm as the bur descends towards the gingival
  • The cut is in one direction only, from outwards to inwards. Never attempt to cut from the inside towards the outside, i.e., the reverse direction. You will damage the adjacent tooth.
  • The inherent safety of the above method is that the direction of cut terminates into a void- the prepared space of the internal box. The only error possible for an erroneous cut is by not controlling the handpiece correctly.
  • Repeat these steps for the other wall of each box: develop each wall similarly –  3 delicate sequential cuts, moving the patient’s head to optimize visualization. Replace the bur if sharpness is declining and hand pressure is  incrementally increasing. Dull burs invite accidents.
  • What has been accomplished?


The method just described is a significant improvement over standard methodology, and correctly and carefully followed, will result in freedom from damage to the adjacent tooth. It’s flaws are that the gingival/proximal walls do not have a radius at the  line angle, and the gingival margin is not beveled.

See Class II Gingival Margin for a method for preparing the gingival margin of Class IIs. Despite the above technique departures from an  ideal outline form, flowable of correct viscosity, heated prior to placement, see Heated Resins and used as the first increment almost always achieves a flawless seal  at this sharp external line angle. A periodontal probe can be used inside the matrix will reliable carry flowable into the external line angle. However, this form does not meet the ideal form to minimize  stress-concentration, which dictates rounded internal and external line angles, so we live at the momentd with a workable compromise.

To overcome this, Class II technique requires a safe-sided ultrasonically-driven diamond with desirable proximal and gingival bevels, and  radiused line angles.Such instrumentation does not exist yet for direct composite. At one time in the early 2000s, the Kavo Sonicflex  developed  heads suitable for:

  • cast gold  gingival margins(60 degree gingival bevel, 100 degree proximal bevel)
  • pressed porcelain and CAD/CAM restorations, (65 degree proximal bevel, 20 degree gingival bevel) .
  • For composite preparations, we require a 96 degree proximal wall bevel and a 102 degree gingival bevel. See Enamel Axiom #5, Proportionate Bevels) . This 102-degree  figure for the gingival bevel is selected by adding a 6 degree minimal bevel to that of the normal apical inclination of enamel at the CEJ, averaging about 6 degrees for most proximal boxes. Placing bevels of these inclinations  assure rod-end bonding on all surfaces of the proximal box as well as radiused external line angles, avoiding unnecessary sacrifice of enamel, and meeting the Enamel Axioms#1  #2 and #3 and #4

In the early 2000s the Kavo website showed EM studies revealing an excellent finish achieved with ultrasonic diamonds. Results were better than those showing cavosurface chipping and irregularity at attained in an EM study of rotary carbide and diamond instrumentation at November 2012 by Clinician’s Report

These tips appear to have been discontinued. We appear to be going backwards. Clearly, we have a distance to go before today’s clinical practices are superceded by better methods.

Go back to class II restorations