Dissolved methane samples from groundwater are prone to error due to degassing during sampling. The context of this potential error is important when using gas concentration data in scientific assessments, e.g. when assessing gas sources and migrations associated with unconventional gas resources. This research aimed to better understand the potential for methane gas loss during sampling. We collected novel experimental data from laboratory experiments to observe the behaviour of methane exsolution due to pressure changes. Analytical and numerical simulations were also used to assess gas loss due to atmospheric exposure and the effect of desorption on total free gas capture during well purging, respectively. The effect of buoyancy on gas migration in an aquifer, well and pump tube was also considered. Our experimental data show that there is an initial loss of dissolved methane due to absolute pressure changes, but the magnitude and rate of this degassing depends on the rate of depressurisation. Previously, the rate of degassing due to changes in depressurisation has not been measured and the seminal data presented here should inform further research. Analytical models show that atmospheric exposure is not a major source of error as significant gas loss via this process is unlikely in the timeframe required to take a dissolved methane sample. Numerical modelling highlighted a number of potential errors associated with closed-system sampling methods that have been largely overlooked to date, including error due to gas desorption. We propose that direct-fill methods be used as the minimum standard for dissolved methane data collection as this method produces consistent error and can be safely and repeatedly used at any well type. A standard plot for presenting dissolved methane data collected via direct-fill methods is proposed. This approach provides a means of directly comparing dissolved methane data sets.