OBJECTIVE: To determine predictors of Global Assessment of Functioning (GAF) score changes in Australian subjects with schizophrenia treated with risperidone long-acting injection (RLAI). METHODS: e-STAR (electronic-Schizophrenia Treatment Adherence Registry) is an ongoing international observational study of schizophrenic patients who commence RLAI. Data collected retrospectively (12 months) and prospectively (2 years) includes: patient demographics, medications, hospitalisations, Clinical Global Impression-Severity (CGI-S), GAF and adverse events. A general linear model (GLM) was used to determine the impact of baseline characteristics on GAF improvement over 12 months. RESULTS: Data from 399 patients from 15 psychiatric services were available for this analysis. The average age was 38.9±13.2 years (mean ± SD), the mean duration of illness was 12.0±9.6 years, and 68% were male. The baseline CGI-S was 4.5±1.0 and GAF was 42.4±14.1. The greatest improvements in GAF score occurred in the first 3-months following switch to RLAI, however, significant improvements also occurred between 3- and 6-months (baseline: 42.8±14.3; 3-months 50.3±15.5; 6-months 53.0±15.3, p<0.001 [0-3 months, and 3-6 months], n=328). Those patients with baseline GAF scores less than 30 were more likely to show greater improvements in their GAF functioning over the 12-month period (GLM coefficient 28.4 [CI:22.2-34.7, p<0.001]), as were patients with baseline GAF scores of 30-60 (GLM coefficient 14.5 [CI:8.8-20.1, p<0.001]). Patients whose last does of RLAI was 25mg and who stayed on RLAI for 12-months were more likely to have positive changes in their GAF over the 12-month period (GLM coefficient 5.3 [CI:1.8-8.8 and 5.0 [CI: 1.7-8.4, p~0.003], respectively). CONCLUSIONS: Patients with a low GAF score at baseline, and those who stayed on 25mg of RLAI are more likely to show positive improvements in GAF functioning when treated over a 12-month period. The most significant improvements in functioning are seen in the first six months of treatment.