Background: Investigations of cognitive deficits after frontal lobe damage have commonly relied on multidimensional tests and relatively coarse specification of lesion anatomy. Some form of impairment in attention is often asserted to cause the revealed deficits. Objective: To describe a disorder of attention in patients with frontal damage using a theoretical model of the fundamental cognitive processes that underlie attention. Methods: The ability to perform a task of concentrated responding was studied in 43 patients with well-defined chronic frontal lesions and 38 control subjects using a continuous reaction time (RT) test. Performance measures were mean RT, RT across blocks of the test, and errors. Lesion measures were coarse localization and a hot-spot analysis to detect finer grained lesion effects. Results: Only patients with lesions in the right superomedial (SM) frontal regions had significantly prolonged RT consistently across the entire test. The critical lesion was in Brodmann’s areas 24, 32, 9, and 46 days and in the adjacent corpus callosum. Patients with lesions in left lateral frontal (LL) regions made significantly more errors on the 20%of trials in the first block. The critical lesion was in areas 44, 45, and 47/12. Conclusion: Concentrating attention to respond is affected by lesions in two different frontal regions for reasons that reflect impairments in different cognitive processes. Right superomedial lesions cause an insufficient energizing of attention to respond. Left lateral lesions cause defective setting of specific stimulus-response contingencies. Constrained tests of attention can demonstrate impairments in specific cognitive operations following lesions to different regions of the frontal lobes. Background: Investigations of cognitive deficits after frontal lobe damage have commonly relied on multidimensional tests and relatively coarse specification of lesion anatomy. Some form of impairment in attention is often asserted to cause the revealed deficits. Objective: To describe a disorder of attention in patients with frontal damage using a theoretical model of the fundamental cognitive processes that underlie attention. Methods: The ability to perform a task of concentrated responding was studied in 43 patients with well-defined chronic frontal lesions and 38 control subjects using a continuous reaction time (RT) test. Performance measures were mean RT, RT across blocks of the test, and errors. Lesion measures were coarse localization and a hot-spot analysis to detect finer grained lesion effects. Results: Only patients with lesions in the right superomedial (SM) frontal regions had significantly prolonged RT consistently across the entire test. The critical lesion was in Brod mann’s areas 24, 32, 9, and 46 days and in the adjacent corpus callosum. Patients with lesions in left lateral frontal (LL) regions made significantly more errors on the 20% of trials in the first block. The critical lesion was in areas 44, 45, and 47/12. Conclusion: Concentrating attention to respond is affected by lesions in two different frontal regions for reasons that reflect impairments in different cognitive processes. Right superomedial lesions cause an insufficient energizing of attention to respond. Left lateral lesions cause defective setting of specific stimulus-response contingencies. Constrained tests of attention can demonstrate impairments in specific cognitive operations following lesions to different regions of the frontal lobes.