Until recently, schizophrenia has proven refractory to molecular genetic investigation, with all advances being hastily followed by retreat. However, more realistic interpretations of old study designs, combined with newer forms of investigation appear at last to be generating reproducible data, suggesting that the field is finally making sustained progress. In this review, we discuss the rationale behind the approaches now in widest application and discuss the results of recent linkage and association studies of schizophrenia. We conclude that, because of advances in the methodology of molecular genetics and improvements in study design, there are solid grounds for believing that susceptibility genes for schizophrenia will be identified in the foreseeable future.