The hydroxyl free radical (OH) is the major oxidizing chemical in the atmosphere destroying about 3.7 Pg of trace gases each year. Global observations of methyl chloroform (CH₃CCl₃), available since 1978 from the ALE/GAGE/AGAGE network, have been widely used to indirectly estimate OH concentrations on hemispheric to global spatial scales and annual time scales using optimal estimation inverse methods and chemical transport models. As reviewed in the IPCC 2007 assessment, these inverse studies suggest that global OH levels grew between 1979 and 1989, then declined between 1989 and 1998, and then recovered again so that in 2003 they are comparable to those in 1979. The circa 1998 OH minimum coincides with major global wildfires and an intense El Nino at this time. The inferred OH levels also exhibit significant inter-annual variations, but only conclusions about their phasing, and not their amplitude, are robust. These decadal and inter-annual OH estimates remain even after accounting for additional recent lingering anthropogenic CH₃CCl₃ emissions, and the proposal that the polar oceans stored methyl chloroform during its pre-peak (1992) years and began re-emitting it in subsequent years. We will update previous OH estimates using AGAGE CH₃CCl₃ measurements through 2007. Looking to the future, the continued use of CH₃CCl₃ for these purposes could potentially become very limited because its concentrations are rapidly decreasing and because any lingering future anthropogenic, wildfire and oceanic emissions would increase the errors in the OH estimates. We will present an extensive analysis of southern hemispheric methyl chloroform measurements that indicates that the proposed substantial oceanic re-emission did not occur. We will also review the significant evidence for rapidly declining anthropogenic emissions and small wildfire emissions. We will finally show the results of a numerical inverse modelling study that calculates the errors in future OH estimates using CH₃CCl₃ under a range of assumptions regarding instrumental precision, atmospheric variability, and lingering emissions of this gas. We conclude that, with achievable increases in instrumental sensitivity and precision, CH₃CCl₃ could remain viable for OH estimations for another decade with an accuracy comparable to that achieved today.