Surface, atmospheric precipitation and deionized water samples were spiked with ??g l-1 concentrations of bromide, and the solutions stored in polyethylene and polytetrafluoroethylene bottles. Bromide was determined periodically for 30 days. Automated fluorescein and ion chromatography methods were used to determine bromide in these prepared samples. Analysis of the data by the paired t-test indicates that the two methods are not significantly different at a probability of 95% for samples containing from 0.015 to 0.5 mg l-1 of bromide. The correlation coefficient for the same sets of paired data is 0.9987. Recovery data, except for the surface water samples to which 0.005 mg l-1 of bromide was added, range from 89 to 112%. There appears to be no loss of bromide from solution in either type of container.Surface, atmospheric precipitation and deionized water samples were spiked with mu g l** minus **1 concentrations of bromide, and the solutions stored in polyethylene and polytetrafluoroethylene bottles. Bromide was determined periodically for 30 days. Automated fluorescein and ion chromatography methods were used to determine bromide in these prepared samples. Analysis of the data by the paired t-test indicates that the two methods are not significantly different at a probability of 95% for samples containing from 0. 015 to 0. 5 mg l** minus **1 of bromide. The correlation coefficient for the same sets of paired data is 0. 9987. Recovery data, except for the surface water samples to which 0. 005 mg l** minus **1 of bromide was added, range from 89 to 112%. Refs.