Spatial working memory was investigated in 20 patients with unilateral neurosurgical excisions of the frontal cortex (UFL), nine with right (RFL) and eleven with left lesions (LFL), comparing their performance to a matched control group. Spatial memory was tested using the Executive Golf Task, a test that also measures spatial strategy formation. Overall the UFL were significantly impaired, the greatest impairment being found in the RFL group. The difference between the RFL and LFL groups was abolished when a measure of strategy formation was used as a covariate in the analysis. A further test of spatial working memory, the Owl Spatial Working Memory Task, which prevents the use of a spatial strategy, showed a significant and equivalent impairment in both the RFL and LFL patients. The data are consistent with neuropsychological and functional neuroimaging investigations supporting the role of the pre-frontal cortex in spatial working memory.