Purpose: To proof the hypothesis that the results achieved by microfracture of the knee in clinical practice are not statistically comparable to those presented in controlled studies. Methods: Our literature search focused on patients with full-thickness cartilage defects of the knee who were treated with microfracture without implantation of a scaffold or injection of substitutes and whose functional capacity was assessed by clinical scores. Whereas these patients formed group A, group B consisted of 26 patients who had been treated at our Level I trauma center. The decision if microfracture was appropriate was made during an arthroscopy of the knee that was scheduled for the repair of meniscal and/or ligamental deficiencies. Due to missing preselection no exclusion criteria were defined. We compared the two groups by performing a correlation analysis. Results: Four randomized controlled trials and one prospective cohort study were detected which provided 187 patients for group A. Surprisingly, a very high correlation (Pearson coefficient r = 0.924, p = 0.008) was calculated between the score values, indicating similar postoperative outcome in both groups. Conclusions: Controlled studies may reflect general population results and therefore provide estimates of treatment effects which are applicable in usual clinical practice.