Background: CLOX, a clock drawing test protocol uniquely sensitive to impairment of executive functions, has been proposed as a screening tool for mild cognitive impairment (MCI), but data about its diagnostic efficiency are lacking. Methods: There are data for 196 subjects, age >= 60 years, referred to a memory clinic for cognitive complaints. After extensive neuropsychological testing, 64 were diagnosed as cognitively normal and 132 with MCI. Results: At standard cutoffs, both CLOX subtests had a fair specificity (CLOX1 72%, CLOX2 92%) but unacceptably low values of sensitivity (CLOX1 54%, CLOX2 28%) and likelihood ratio (CLOX1 1.91, CLOX2 3.59) for MCI. The use of different cutoffs or the combination of CLOX with the Mini-Mental State Examination (MMSE) did not statistically increase diagnostic efficiency. Conclusion: CLOX, either alone or in combination with MMSE, is not a useful screening test for MCI in a clinical setting. Copyright (C) 2010 S. Karger AG, Basel