Background: The patient’s preference for treatment is dynamic across different cultures and at various stages of cancer. Yet little is known regarding their priorities in making treatment decisions. The study's primary objective is to examine the preference weightings in terms of risk/ benefit attributes of therapy in metastatic prostate cancer (mPC) patients in both the hormone-sensitive (mHSPC) and castration-resistant (mCRPC) settings. Methods: This is a non-interventional cross-sectional survey study. A discrete choice experiment was used to assess their priorities on treatment options. 300 consecutive patients with mPC (200 mHSPC and 100 mCRPC) diagnosed within 5 years were recruited from the specialty clinics of our hospital from Feb 2023 to Jul 2023. A facilitated group interview session of 5 patients was conducted to evaluate the importance and relevance of potential attributes selected from the literature search. Five attributes were adopted, and each has 3 levels compromised between real-life rates and expectations of patients from the interview session. Patients were randomized into 2 blocks by recruitment sequence and chose 1 medication profile out of 2 throughout the 9 questions in each block of the questionnaire. A mixed logit model was used for data analysis. Results: Median age was 75 (IQR:71-81). One hundred and seventy (56.7%) had no income, and 245 (81.7%) claimed to take care of themselves. The mean maximum amount affordable for treatment was US$20,456 (SD:43,568), and 160 (53.3%) said they would not consider further treatment if the cost exceeded their affordability. Compared to intravenous formulation, oral medication once/day had a non-statistically significant favorable preference (0.06 p:0.272), and oral medication twice/day had a negative statistically significant preference (-0.20 p:0.040). The percentage of retained self-care ability (4.37 p<0.001) and percentage of extension in life expectancy (2.83 p<0.001) had a statistically significant favorable preference. In contrast, the chance of adverse effects leading to complicated medication treatment or hospitalization (-6.97, p<0.001) and treatment cost (in HK$ million) (-3.14 p<0.001) had a negative statistically significant preference. The mCPRC group was more sensitive to changes in treatment cost (-3.61 vs. -2.97), percentage of extension in life expectancy (3.47 vs. 2.55) and the chance of adverse effects (-7.55 vs -6.80) than the mHSPC group. Those who could afford treatment were more sensitive to the percentage of retained self-care ability (4.89 vs. 4.02) and the chance of adverse effects (-7.57 vs -6.70) than those who could not. Conclusions: The chance of adverse effects took the largest weighting in treatment decisions, followed by the percentage of retained self-care ability.