Objective The efficacy of anamorelin in pancreatic cancer (PC) patients with a poor performance status (PS) is uncertain, as previous trials have excluded such patients. This study evaluated the efficacy of anamorelin in PC patients with a poor PS (2) compared with those with a good PS (0-1). Methods We retrospectively reviewed consecutive PC patients with cachexia who received anamorelin at our institution. The primary outcome was the proportion of responders, defined as those who maintained or gained body weight and appetite over 12 weeks. The secondary outcomes included anamorelin treatment duration, proportion of patients who discontinued anamorelin within 4 weeks (early discontinuation), and the overall survival. Results Forty-five patients (35/10) were included in this study. The proportion of responders was significantly lower in patients with a poor PS than in those with a good PS (0% vs. 37%, p=0.042). Moderate weight loss (5-10%) and administration of pancreatic enzyme replacement therapy were associated with a response to anamorelin. A poor PS was significantly associated with a shorter treatment duration of anamorelin (14 vs. 93 days, p<
0.001), a higher proportion of patients who discontinued anamorelin within 4 weeks (70% vs. 17%, p=0.003), and a reduced survival (62 vs. 188 days, p<
0.001). A poor PS was associated with early discontinuation of anamorelin. Conclusion The efficacy of anamorelin is extremely limited in PC patients with a poor PS. Patients with PC with a poor PS may not be good candidates for anamorelin compared to those with a good PS.