Gingival recessions may be classified according to a number of different systems. One of the most frequently used systems both in the literature and everyday practice is that proposed by Preston D. Miller in 1985. This classification assigned recessions to four different classes and gave a prognosis with regard to complete root coverage for each of the four situations following surgical treatment. According to Miller himself, complete coverage may be predicted in both Class I and II recessions, whereas its attainment is unpredictable in Class III and impossible in Class IV. Given that complete root coverage is attained in some Class III recessions, an investigation was planned to determine the presurgical variables that could help predict these results. Presurgical, surgical, and postsurgical conditions were studied for 121 Class III recessions in 50 patients, of which 47% achieved complete root coverage. Following statistical analysis using linear regression techniques, it was observed that the outweighing variables that explained these results were: soft tissue interproximal integrity, the use of grafts with a thickness of greater than 2 mm, interproximal bone loss not exceeding 3 mm, and an initial recession width not greater than 3 mm. Complete root coverage is possible in some Class III recessions. The Miller classification could be extended if these findings are confirmed by further research.