To assess whether patient-controlled admissions (PCA) to psychiatric hospital wards have an effect on total health care costs in the short term and longer term compared with patients receiving treatment as usual (TAU). Based on Danish register data and using coarsened exact matching, patients who signed a contract for PCA during 2013-2016 were matched on observables with a control group of TAU patients identified from psychiatric registers. As primary outcome patients' expenditure was explored in 6-monthly intervals over the first 12 months, and as secondary outcome over the 12-24 month period after index month. To control for unobservable selection a quasi-experimental difference-in-difference design was applied on the matched population using the principle of intention to treat. 414 PCA patients (out of 423 available) were successfully matched with 7181 never exposed TAU patients (out of 67,331 available). Total healthcare costs for PCA patients were €8887 (95% CI 708-17,067) higher per month than unexposed TAU patients in the first six months after index. Monthly psychiatric costs were €8922 (95% CI 708-17,067) higher for PCA patients than for TAU patients. These differences persisted, in the next three six-monthly periods, albeit were not statistically significant, in the second six-monthly period after index. In conclusion, driven by the psychiatric costs, implementing PCA increased the total healthcare costs in the short term compared with TAU. More research of the impact on health care costs in a broader and long-term perspective is still needed.