The objective was to demonstrate the potential of axial T2 mapping for quantification of untreated early-stage patellar cartilage lesions over time and to assess its capability as a potential predictive marker for future progression. Thirty patients (mean age, 36.7±11.1 years; 16 males), with early-stage patellar cartilage defects (≤ICRS grade 2) at baseline and no treatment during follow up (4.0±1.6 years) were enrolled. Morphological cartilage changes over time were subdivided into a Progression, Non-Progression Group and Regression Group. Quantitative analysis of cartilage defects and healthy reference was performed by means of global and zonal T2 mapping (deep and superficial cartilage T2 values) at both time points. Statistical evaluation included analysis of variance (ANOVA), paired t Test's and ROC analysis. The Progression Group (N=11) had significantly higher global T2 values at baseline (57.4±7.8ms) than patients without (N=17) (40.6±6.9ms) (P<0.01). Furthermore the Non-Progression Group showed only a minor increase in global T2 relaxation times to 43.1±7.9ms (P=0.07) at follow up, whereas in the progression group global (68,7±19ms: P=0.02) and superficial T2 values (65,8±8.2-79.8±24.4ms; P=0.03) increased significantly. T2 values for healthy reference cartilage remained stable. In 2 patients an improvement in ICRS grading was observed (Regression Group) with decreasing T2 values. The ROC analysis showed an area under the curve of 0.92 (95%CI 0.82-1.0). At a cut-off value of 47.15ms, we found a sensitivity of 92% (false-positive rate of 18%) for future progression of cartilage defects. This study provides evidence regarding the possible potential of axial T2 mapping as a tool for quantification and prediction of patellar cartilage defect progression in untreated defects.