OBJECTIVE Low volume sweat samples are considered unreliable for the diagnosis of cystic fibrosis, based on the assertion that sweat conductivity and chloride are reduced at lower sweating rates. We aimed to re-evaluate the relationship between sweat volume and test results. METHODS We reviewed all sweat tests performed in our institution to assess the relationship between sweat volume and conductivity, and between sweat volume and sweat chloride. We also compared results between pairs of sweat tests taken simultaneously from a single patient, one with sweat volume below and the other above the currently accepted minimum volume (15 microl). RESULTS A weak inverse relationship between sweat volume and sweat conductivity was found (n=1500, R2=0.105, p<0.001). There was no correlation between sweat volume and sweat chloride (n=463, R2=0.002, p>0.05). In discordant pairs (one below and one exceeding the accepted minimum volume), the mean test result in the low volume sample was slightly higher than its counterpart. In 76 such pairs, mean conductivity was 41.1+/-14.6 mmol/l in the lower volume sample, compared with 36.8+/-16.0 mmol/l in the higher volume sample (p<0.001). Similarly, in 33 of the pairs, mean sweat chloride was 28.4+/-15.7 mmol/l in the lower volume sample compared with 25.1+/-15.2 mmol/l in the higher volume sample (p=0.004). CONCLUSIONS A normal sweat conductivity and/or chloride value from a sweat volume <15 microl in a patient whose clinical symptoms are not very suggestive of cystic fibrosis, renders this diagnosis unlikely. In contrast, elevated sweat chloride or conductivity measured from a sample whose volume is <15 microl may represent an artefact related to the low volume.