ObjectiveEvidence-based personality disorder (PD) treatments are dominated by interventions targeting Borderline PD, although clinical populations characteristically include different PD features and severity. Personality functioning is a new concept intended to capture common features across PDs. This study aimed to investigate longitudinal improvement of personality functioning in a clinical sample assigned to PD treatment.MethodAn observational, large, longitudinal study of patients in PD treatments on specialist mental health service levels (N = 1,051). DSM-5 PDs were systematically assessed on referral. Personality functioning was repeatedly assessed (LPFS-BF-2.0), supplemented by symptom distress (anxiety: PHQ-GAD-7, depression: PHQ-9), and social/occupational activity (WSAS, work/study activity). Statistics were linear mixed models.ResultsThirty per cent had personality difficulties below PD threshold. Among PDs, 31% had Borderline (BPD), 39% Avoidant (AvPD), 15% not otherwise specified, 15% other PDs, and 24% > one PD. More severe initial LPFS-BF was associated with younger age, presence of PD and increasing number of total PD criteria. Across PD conditions, LPFS-BF, PHQ-9 and GAD-7 improved significantly (overall effect size 0.9). Mean duration of PD treatment was 15 (SD 9) months. Drop-out rates were low (12%). LPFS-BF improvement-rates were higher for BPD. Younger age was moderately associated with slower PHQ-9 improvement. Work/study activity was initially poor, poorer levels associated with AvPD and younger age, and improvement was non-significant across PD conditions. AvPD was associated with slower WSAS improvement-rates.ConclusionPersonality functioning improved across PD conditions. The results highlight BPD improvements. The study points to challenges concerning AvPD treatment, poor occupational activity and age-related differences.