The purpose of this study was to examine sex-stratified independent predictors of patient-initiated discharge from an inpatient withdrawal management service and to determine whether those predictors differed by sex. This study compared people who had self-initiated versus planned discharges and used sex-stratified generalized estimating equations models to identify independent predictors of patient-initiated discharge. Predictors examined included age, ethnicity, substance of concern, tobacco use, mental health comorbidities, day of discharge, referral source, children, and social assistance funds. Among females, there were 722 discharges, 116 of which were patient initiated. Among females, increasing age was associated with lower odds of patient-initiated discharge (OR = 0.97, 95% CI [0.95, 0.98]). Racialized females were nearly 2 times more likely to experience patient-initiated discharge compared with White females (OR = 1.8, 95% CI [1.09, 3.00]). Compared with weekdays, weekends were associated with over 4 times the odds of patient-initiated discharge (OR = 4.77, 95% CI [2.66, 8.56]). Having one or more mental health comorbidities was associated with lower odds of patient-initiated discharge compared with having no mental health comorbidities (OR = 0.51, 95% CI [0.32, 0.82]). Among males, there were 1,244 discharges, 185 of which were patient initiated. Among males, increasing age was associated with decreased odds of patient-initiated discharge (OR = 0.97, 95% CI [0.95, 0.98]). Compared with weekdays, weekends were associated with nearly 15 times the odds of patient-initiated discharge (OR = 14.9, 95% CI [9.11, 24.3]). Males and females have shared and unique predictors of patient-initiated discharge. Future studies should continue to examine the influence of sex and gender on engagement with addictions care.