Mental disorders affect 10-20 % of the young population in the world. Major depressive disorder (MDD) is a common mental disease with a multifactorial and not clearly explained pathophysiology. Many cases remain undetected and untreated, which influences patients’ physical and mental health and their quality of life also in adulthood. The aim of our pilot study was to assess the prediction value of selected potential biomarkers, including blood cell counts, blood cell ratios, and parameters like peroxiredoxin 1 (PRDX1), tenascin C (TNC) and type IV collagen (COL4) between depressive pediatric patients and healthy peers and to evaluate a short effect of antidepressant treatment. In this study, 27 young depressive patients and 26 non-depressed age-matched controls were included. Blood analyses and immunological assays using commercial kits were performed. Platelet count was the only blood parameter for which the case/control status was statistically significant (p=0.01) in a regression model controlling for the age and gender differences. The results from ELISA analyses showed that the case/control status is a significant predictor of the parameters PRDX1 (p=0.05) and COL4 (p=0.009) in respective regression model considering the age and gender differences between MDD patients and controls. A major finding of this study is that values of platelet count, monocyte to lymphocyte ratio, white blood cell, and monocyte counts were assessed by the Random Forest machine learning algorithm as relevant predictors for discrimination between MDD patients and healthy controls with a power of prediction AUC=0.749.