The frequency of patients with presumptive tuberculosis (TB) who are not investigated by sputum smear microscopy is unknown in Pakistan. Using a simple intervention comparing patient and laboratory registers, patients with presumptive TB were identified in two districts from July to December 2013, a list of missing patients was prepared and the patients traced. The intervention significantly reduced the number of patients with presumptive TB lost, from 8.5% before the intervention to 6.9% after. A systematic comparison of out-patient and laboratory registers, followed by tracing missing persons, can reduce the proportion of patients with presumptive TB lost before diagnosis.