OBJECTIVES To determine the cost-effectiveness of atypical long-acting injectable (LAI) antipsychotics in treating relapsed chronic schizophrenia from the viewpoint of the Finnish National Health Service. METHODS A 1-year decision tree was adapted for use with patients in relapse, guided by an expert panel. Drugs included available atypical long-acting antipsychotics: paliperidone (PP-LAI), risperidone (RIS-LAI), olanzapine (OLZ-LAI) and aripiprazole (ARI-LAI). Rates of adherence, success, relapse and hospitalization were taken from the literature. Prices were obtained from standard lists and expressed in 2014 euros: drugs, psychiatrists/physicians, psychiatric nurse, inpatient and outpatient hospital care. Outcomes included expected cost/patient treated, QALYs, rates of re-hospitalization, emergency room (ER) visits and days in relapse. The primary analysis was the incremental cost per QALY. These preliminary results were tested with 1-way sensitivity analyses on important inputs.  RESULTS Over the 1-year time horizon, PP-LAI had the lowest total cost of 34,446€ per patient, RIS-LAI cost 37,338€, ARI-LAI cost 37,433€ and OLZ-LAI cost 41,384€. PP-LAI had the highest number of QALYs (0.686), followed by OLZ-LAI with 0.680, RIS-LAI with 0.674 and ARI-LAI with 0.671. PP-LAI also had the lowest rates of all negative outcomes. Re-hospitalization rates were 10.1%, 12.5%, 12.4% and 12.2% for PP-LAI, RIS-LAI, ARI-LAI and OLZ-LAI, respectively, Respective ER visits were 20.6%, 23.2%, 25.4% and 21.7%. Patients receiving PP-LAI experienced 224.0 relapse days, as opposed to 234.8 with RIS-LAI, 237.5 with ARI-LAI and 234.8 with OLZ-LAI. In 1-way sensitivity analyses, costs were robust (required changes >20%) against changes in drug price, primary rate of success, rates of relapse, dropouts and adherence.  The driver of the model for all drugs was hospitalization, comprising from 70%-81% of the total cost; drug costs constituted 12%-22% and medical care 8%-15%. CONCLUSIONS PP-LAI was shown to have the lowest cost and best clinical outcomes, and hence should be the atypical LAI of choice.