There are several methodological difficulties to address when evaluating acute psychiatric services. This study explored potential methods in evaluating the effectiveness of women's crisis houses compared with psychiatric wards in a pilot patient preference randomized controlled trial. Women requiring voluntary admission to a psychiatric hospital or women's crisis house were asked to enter this pilot and different options for recruitment were explored, including different recruitment sites in the pathway to admission and methods for including women without capacity. Forty-one percent (n = 42) of women entering the study agreed to be randomized and 59% (n = 61) entered patient preference arms. Only 7% of women were recruited before admission and 1 woman without capacity entered the study, despite procedures to facilitate this. Recruitment of patients with acute psychiatric crises is therefore challenging; researchers evaluating acute services should establish a consensus on how ethically and practically to recruit patients in this setting.