OBJECTIVES: Trials of new oncology treatments often allow patients to crossover from control to experimental treatment either at disease progression or a specific time point. If crossover occurs after progression, progression free survival is largely unaffected; however overall survival (OS) is confounded in the control arm.  Patients switching treatments often have different prognoses, resulting in a biased estimated OS difference. Advanced statistical methods to adjust for crossover include Rank Preserving Structural Failure Time (RPSFT) models, Inverse Probability of Censoring Weights (IPCW), and Iterative Parameter Estimation (IPE). To determine acceptance of these techniques in Australia, Public Summary Documents (PSDs) reporting on the Pharmaceutical Benefits Advisory Committee's (PBAC’s) decision-making in oncology were reviewed. METHODS: Oncology PSDs were examined, and the method of adjustment (if any), PBAC concerns and outcome were assessed. RESULTS: Thirty-one submissions included trials allowing crossover; 16 (52%) presented adjusted estimates of OS.  Common indications were: renal cell carcinoma, non-small cell lung cancer (NSCLC), melanoma, colorectal cancer (CRC), and pancreatic neuroendocrine tumour.  The most common method was RPSFT (11 products), followed by IPCW (4 products).  IPE, marginal structural modelling, and a two-stage Weibull approach were each reported once.  The RPSFT approach appeared to raise fewer concerns than the IPCW, usually because the IPCW correction was considered less reliable when a high proportion of patients crossed-over.  In NSCLC and CRC, the PBAC considered unadjusted crossover to be appropriate and reflect a relevant comparison between first- and second-line therapy.  Only five PSDs reported results of more than one method. The PBAC expressed a preference to see a range of approaches and a clearly justified selection of the most appropriate method. CONCLUSIONS: A range of adjustment techniques were used to support submissions in Australia, with RPSFT being most common.  It is important to clearly justify the need for adjustment and the selection of the most appropriate technique.