Despite the preponderance of treatment outcome predictors in patients with borderline personality disorder (BPD), the predictive value of measures of impulsiveness is inconclusive. This naturalistic study consecutively included hospitalized patients with BPD (N = 99) who underwent a standardized and structured 12‐week inpatient treatment programme, which integrated cognitive‐behavioural and psychodynamic elements. The Brief Symptom Checklist (BSCL) was applied as outcome measure over four time points: pretreatment, posttreatment, first follow‐up at 6 to 8 weeks and second follow‐up at 1 year after discharge. Impulsiveness was measured using the Barratt Impulsiveness Scale (BIS) at the pretreatment time point. The BSCL significantly decreased between pretreatment and posttreatment, followed by an increase after posttreatment without reaching pretreatment extent. The temporal course of the BSCL significantly varied with pretreatment BIS in that patients with higher impulsiveness revealed a stronger re‐increase of symptom severity from posttreatment to end of follow‐up than those with lower impulsiveness. The least impulsive patients thereby showed no rebound effect. The robustness of the results was examined by cross‐validation. The results indicate that irrespective of the level of impulsiveness, patients with BPD profit from a structured inpatient treatment. However, long‐term treatment success was impaired in patients with high level of impulsiveness at pretreatment. Thus, self‐ratings of impulsiveness in BPD patients can be utilized for treatment planning. After discontinuation of interventions, relapse prevention should be implemented early in high impulsive patients as symptoms recrudesce in the course after discharge.