There is a need for more objective outcome measures for chronic cough. In this pilot study we sought to investigate the utility of the mannitol challenge as a cough-provocation test in non-asthmatic chronic cough. We studied 16 healthy controls and 13 subjects with chronic cough. We assessed cough severity using a visual analogue score, capsaicin cough sensitivity, health status using the Leicester Cough Questionnaire and the dose of mannitol to cause 2 (C2) or 5 (C5) coughs. In all of the subjects with chronic cough and 6 of the controls we assessed the 1-week repeatability of the mannitol challenge. We found that in those subjects with chronic cough the geometric mean (logSEM) mannitol C2 and C5 was heightened compared to controls (C2: 4 (0.2) versus 16 (0.1); p = 0.04 and C5: 63 (0.1) versus 251 (0.1); p = 0.04). Cough visual analogue score, capsacin-induced cough sensitivity and health status were also altered in chronic cough compared to healthy controls, but in those subjects with chronic cough none of these outcomes was correlated with the mannitol C2 or C5. The repeatability of the mannitol challenge assessed by intraclass correlation was C2 = 0.53 and C5 = 0.59. A cut-off in the dose of mannitol of 62 mg/ml for C2 and 550 mg/ml for C5 had a sensitivity of 69 and 62% and specificity of 69 and 81% respectively to distinguish chronic coughers from healthy controls. In conclusion, the mannitol challenge my have potential as a novel cough challenge test and further work is required to extend our findings and to assess whether it has utility in different causes of chronic cough.