Purpose We investigated the added value of difusion-weighted imaging (DWI)/apparent difusion coefcient (ADC) in the categorization of small hepatic observation (≤20 mm) detected in patients with chronic liver disease in reference to LI-RADS (liver imaging reporting and data system) classifcation system. Methods We prospectively evaluated 165 patients with chronic liver disease with small hepatic observations (≤20 mm) which were previously categorized as LI-RADS grade 3–5 on dynamic contrast-enhanced CT (DCE-CT). All patients were submitted to a functional MRI including DCE and DWI. Using LI-RADS v2017, two radiologists independently evaluated the observations and assigned a LI-RADS category to each observation using DCE-MRI alone and combined DCE-MRI and DWI/ADC. In the combined technique, the radiologists assigned a LI-RADS category based on a modifed LI-RADS criteria in which restricted difusion on DWI was considered a major feature of HCC. We evaluated the inter-reader agreement with Kappa statistics and compared the diagnostic performance of the LI-RADS with two imaging techniques by Fisher’s exact test using histopathology as the reference standard. Results Combined technique in LI-RADS yielded better sensitivities (reader 1, 97% [65/67]; reader 2, 95.5% [64/67]) for HCC diagnosis than DCE-MRI alone (reader 1, 80.6% [54/67], p=0.005; reader 2, 83.6% [56/67], p=0.04). The specifcities were insignifcantly lower in combined technique (reader 1, 88.4% [107/121]; reader 2, 77.7% [94/121]) than in DCE-MRI alone (reader 1, 90.9% [110/121], p=0.67; reader 2, 79.3% [96/121], p=0.88). The inter-reader agreement of the LI-RADS scores between combined technique and DCE-MRI was good (κ=0.765). Conclusion The use of DWI/ADC as an additional major criterion, improved the sensitivity of LI-RADS in the diagnosis of HCC while keeping high specifcity.