This cross-sectional study assessed the potential of colored periodontal probes (CPP) to classify gingival phenotype in terms of gingival thickness (GT). Methods Buccal GT in 3 anterior teeth in each of 50 patients was measured by transgingival sounding and classified by 3 different methods by 8 examiners. Specifically, the diagnostic potential of visual judgement and transparency of a standard periodontal probe (SPP) to discriminate thin and thick gingiva, and of CPP to discriminate thin, medium, thick, or very thick gingiva was assessed. Results GT ranged from 0.57-2.37mm. Using CPP resulted in a medium judgement in 87 on average, and only between 1-10 cases/examiner were judged as thick or very thick. Considering 1mm GT as relevant cut-off value, all methods showed a high positive predictive value (≥0.82) to identify thick cases, but also a high false omission rate (≥0.73) indicating that many cases classified as thin were actually thick. Further, 881mm, were not classified as thin with CPP; this was inferior to SPP, for which, however, still 641mm thick were wrongly classified. The highest, yet moderate agreement among examiners was achieved by SPP (κ = 0.427), while visual judgement and CPP showed only fair (κ = 0.211) and slight agreement (κ = 0.112), respectively. Conclusion Using CPP resulted in most of the cases in a medium judgement. It seems that CPP cannot distinctly discriminate between “thick” and “very thick” cases and fails to capture the thin high-risk cases.