There has been no attempt to explicate elderly patients' preferences during nurse-patient interaction in the context of a prison setting. A purposive sample of 347 was chosen from a national Philippine prison based on the following inclusion criteria: (a) 55 years old and above, (b) incarcerated, (c) no psychological impairment, and (d) can amply read and write. Via conjoint analysis (CA), 36 from 81 nurse-patient interaction cards were produced according to orthogonal plans, as specified by SPSS. The conjoint analysis model proved to be fit: Pearson R=.993, (p<.05) Kendall's tau=.944 (p<.05), Kendall's tau for holdouts=.799 (p<.05). The use of CA showed the top 3 attributes: language (21.26%), smile (18.49%), and eye contact (14.19%). Overall, the following levels are highly preferred by respondents: "vernacular'' language (utility=1.377, SE=.21.26), "with'' smile (utility=1.198, SE=.082), and "constant'' eye-contact=gaze (utility=.919, SE=.082). Significant differences in nurse-patient interaction were noted in age brackets, current health condition, and years of incarceration, except in educational attainment. Overall, geriatric incarcerated individuals preferred nurses who verbally utilize close-ended, nonfrequent, and without follow-up questions, speak vernacular in a low and constant intonation, give encouragement, and praise. When touch is needed, the respondents preferred same gender nurses. Respondents find it more comfortable when nurses utilize 'friendly' space during the interaction, smile, and maintain constant eye contact. To achieve more consistent care, nursing actions must be in accordance with these preferences.