Purpose: Ovarian tumours are the second most common cause of death from gynaecological cancer. There are three types of ovarian cancer based on histopathological examination: benign, borderline, and malignant. However, it is dif-ficult to distinguish the borderline and malignant tumours. Several studies used the apparent diffusion coefficient value to distinguish the ovarian tumour types, with various results. This preliminary report focused more on the use of the minimal ADC (mADC) value on the solid component, to differentiate borderline and malignant ovarian tumours. Material and methods: In 21 cases of borderline ovarian tumours, of which 11 were regarded as malignant and 10 were regarded as borderline following histopathological examination, the mADC value was measured by two different radiologists by using free-hand technique. The intraclass correlation coefficient (ICC) was used to measure the reliability and agreement between the two radiologists. Receiver-operating characteristic (ROC) curves were then calculated to determine the optimum cut-off point. Results: There were statistically significant (p = 0.001) of the mADC value between the borderline and malignant tumours. The intraclass correlation coefficient value showed excellent reliability and agreement between the exam-iners. The ROC curve showed the optimum cut-off point at 0.628 × 10–3 mm2/s (p = 0.001), which yielded 100% sensitivity and 80% specificity. Conclusions: The use of free-hand technique to measure the mADC value on the solid component can be valuable in differentiating borderline and malignant ovarian epithelial tumours. This result may assist clinicians in considering further treatment approaches. © Pol J Radiol 2020.