PET/CT-based Deauville scoring (DS) is routinely used for lymphoma response assessment. However, pathological correlation of DS is not yet precisely documented. In the present study we aimed to pathological confirm the PET/CT-based Deauville scoring (DS) in lymphoma after first-line chemotherapy. Participants undergoing PET/CT for response assessment following first-line treatment were recruited prospectively. DS ≥ 4 lesions were interpreted as PET-positive, while DS ≤ 3 as PET-negative. Participants with a PET-positive lesion or suspicious of inadequate response (DS ≤ 3) were recruited for metabolic core-needle biopsy. True-negative and benign histopathology were kept on follow-up for three months. Histopathological, clinical and imaging findings were assessed for diagnostic performance. Procedure-related complications were also noted. In all, 148/480 participants were PET-positive, and 332/480 were PET-negative. 138/148 PET-positive and 12/332 PET-negative lesions were recruited for biopsy. Biopsy was performed in 147/150 participants (PET-positive 135; PET-negative 12). Three patients with inaccessible lesions were excluded. The diagnostic yield of the procedure was 97.3% (143/147). Histology revealed lymphoma in 106 participants (including 70% of total DS-4, 100% of DS-5a and 73.9% of DS-5b lesions), with three false-negative lesions. DS ≤ 3 lesions were true-negative except one diagnosed with lymphoma (8.3%) on follow-up. Non-lymphomatous malignancies (n = 5), granulomas (n = 12), non-specific inflammation (n = 9) and no residual disease (n = 11) were diagnosed in the rest. No major procedure-related adverse event was noted. A DS-5a lesion suggests residual disease; hence, a biopsy can be prevented unless Richter's transformation is suspected. DS-4 and DS-5b lesions require a biopsy before changing the treatment plan, as a certain number of participants had non-lymphomatous F-18 FDG-avid lesions.