The current trial assessed whether the addition of cisplatin and capecitabine to the nab-paclitaxel–gemcitabine backbone is feasible and active against borderline and locally advanced pancreatic adenocarcinoma (PDAC). Fifty-four chemo-naive patients, aged between 18 and 75 years, with a pathological diagnosis of locally advanced or borderline resectable PDAC were randomised to receive either nab-paclitaxel, gemcitabine, cisplatin and oral capecitabine (PAXG; arm A; N = 26) or nab-paclitaxel followed by gemcitabine (AG; arm B; N = 28). The primary end-point was the tumour resection rate. If at least four such resections were performed, the treatment was considered as active. The secondary end-points were progression-free survival (PFS), overall survival (OS), Response Evaluation Criteria in Solid Tumours response rate, Hartman's pathologic response, carbohydrate antigen 19.9 response rate and toxicity. Eight patients (31%) in the PAXG arm and nine (32%) in the AG arm underwent resection. PFS at 1-year was 58% in arm A and 39% in arm B. OS at 18-month was 69% in arm A and 54% in arm B. In this phase II study, the addition of cisplatin and capecitabine to the AG backbone was feasible and yielded promising results in terms of disease control without detrimental impact on tolerability. The approach warrants further investigation in a phase III study. NCT01730222. •No standard therapy has been yet identified in borderline/locally advanced disease.•A very limited number of randomised trials have been completed in this setting.•This current trial evaluates the activity and feasibility of cisplatin, nab-paclitaxel, capecitabine, gemcitabine (PAXG) and nab-paclitaxel, gemcitabine (AG) regimens.•Both regimens were effective in terms of disease control.•PAXG regimen appears more convenient from several perspectives than AG.