On January 21, 2014, Roche announced that two phase 3 trials of bitopertin for negative symptoms failed to achieve primary end points. An editorial notes that more information is awaited on these negative studies to determine whether the failure to replicate the positive findings in the proof of concept study can be understood by aspects of study design or implementation or by participant heterogeneity. It adds that advances in basic neuroscience and drug development must be matched by improved clinical trial methods including identification of biomarkers to reduce the problem of patient heterogeneity, otherwise, promising new agents associated with moderate therapeutic effects may be abandoned despite the potential for large and consistent effects if a personalised medicine approach can be achieved.