Previous research on semantic priming in schizophrenia has produced conflicting findings. While several studies provided evidence for an enhanced cognitive spreading of activation in schizophrenia patients with formal thought disorder, other research has failed to replicate. The aim of the present study was to resolve some of the ambiguities in the literature. Thirty-two schizophrenic patients (12 with and 20 without symptoms of formal thought disorder according to the PANADSS) and 65 healthy controls were compared in a semantic priming task using word pronunciation. Irrespective of baseline condition (neutral or unrelated condition) patients with formal thought disorder (TD) exhibited a significantly greater indirect semantic priming effect than both non-TD patients and healthy controls. Known confounding variables such as length of illness, neuroleptic dosage and psychomotor slowness did not moderate priming. Results further strengthen the spreading activation model of formal thought disorder put forward by Maher/Manschreck and Spitzer. Data indicate that hyper-priming is not confined to lexical decision tasks. Possible reasons why several studies have failed to replicate greater priming in TD schizophrenic patients are discussed.