BPD is a severe and chronic psychological disorder associated with high rates of completed suicide and enduring disability. Standard DBT is an evidence-based, multi-modal psychotherapy for BPD which includes individual psychotherapy, group skills training, between-session coaching, and therapist consultation. DBT is effective for reducing many difficulties associated with BPD and can also transfer successfully into the community, yet important practice-relevant questions remain. The work in this dissertation involves four separate studies relating to ways DBT can be delivered in the community and exploring who is likely to respond to DBT. The studies draw on treatment outcome data from a total of 140 patients with a diagnosis of BPD collected over several years by community-based DBT teams involved in a multi-site, practice-based research network (PBRN). In Chapter 2, encouraging outcomes are reported from a small evaluation of standard DBT delivered to a cohort of 18-25 year olds with a diagnosis of BPD in a new young adult only DBT programme. Chapter 3 describes an expanded investigation of this young adult only DBT programme, mainstreamed as an early intervention initiative for 18-25 year olds presenting to community services with a diagnosis of BPD. Better outcomes are reported for the young adult programme compared to similar aged young adults in general adult DBT, suggesting advantages for this age-specific mode of delivery, possibly due to group cohesion. In Chapter 4, largely similar outcomes at six months are found for a rationalised, standalone group skills training adaptation of DBT when compared to standard (i.e. all modes) DBT, with some unexpected advantages for standalone group skills on hopelessness and difficulties in emotion regulation. Treatment conditions are not equivalent due to non-random allocation and the exclusion of higher risk patients from standalone skills, but findings offer support for the usefulness of standalone skills among lower-risk patients with a diagnosis of BPD who are willing to accept a group-only intervention when delivered by experienced DBT therapists. In Chapter 5, patient variables are investigated as predictors of outcome following one year of standard DBT for BPD. Gender, employment status, post-traumatic stress disorder (PTSD), and baseline suicide ideation are reported to be associated with change or recovery on borderline symptom severity and general psychopathology. Overall limitations of the dissertation, such as measurement problems, failure to formally assess treatment fidelity, and the inclusion of only completers in the analysis, are discussed in Chapter 6. In addition, several areas for future research are identified: the benefits of PBRNs, group cohesion as a factor in outcomes for DBT, and further study of young adult DBT under randomised trial conditions with longer term and wider follow-up.