A recent report of widespread declines in soil organic C (SOC) in the UK over the 10–25 years until the early 2000s has focussed attention on the importance of resampling previously characterized sites to assess long-term trends in SOC contents and the importance of soils as a potentially volatile and globally significant reservoir of terrestrial C. We have used two sets of long-term experimental plots which have been under constant and known management for over a century and for which historical data exist that allow comparison over recent decades to determine what, if any, changes in SOC content have occurred. The plots used are the Palace Leas (PL) Meadow Hay Plots in north-east England (UK) established in 1897, and from the Park Grass (PG) Continuous Hay experiment established in 1856 at Rothamsted in south-east England. Collectively, these plots represent the only grassland sites in the UK under long-term management where changes in SOC over several decades can be assessed, and are probably unique in the world. The plots have received different manure and fertilizer treatment and have been under known management for at least 100 years. In 1982, total SOC contents were determined for the 0–27 cm layer of six of the PL plots using measurements of SOC concentrations, bulk density and soil depth. In 2006, the same six PL plots were resampled and SOC contents determined again. Four of the plots showed no net change in SOC content, but two plots showed net loss of SOC of 15% and 17% (amounting to decreases of 18 and 15 t C ha−1) since 1982. However, these differences in total SOC content were in a similar range to the variations in bulk density (6–31%) with changing soil water content. In 1959, the soil masses and SOC concentrations to 23 cm depth were measured on six PG plots with fertilizer and manure treatments corresponding closely with those measured on PL. In 2002, the SOC concentrations on the same plots were measured again. On three of the PG plots, SOC concentrations had declined by 2–10%, but in the other three it had increased by 4–8% between 1959 and 2002. If it is assumed that the soil bulk density had not changed over this period, the losses of SOC from the top soils ranged range from 10 to 3 t C ha−1, while the gains ranged from 4 to 7 t C ha−1. When the differences with time in SOC contents for the six PL and the six PG plots were examined using paired t-tests, that is, regarding the plots as two sets of six replicate permanent grasslands, there were no significant differences between 1982 and 2006 for the PL plots or between 1959 and 2002 for the PG plots. Thus, these independent observations on similar plots at PL and PG indicate there has been no consistent decrease in SOC stocks in surface soils under old, permanent grassland in England in recent decades, even though meteorological records for both sites indicate significant warming of the soil and air between 1980 and 2000. Because the potential influences of changes in management or land use have been definitively excluded, and measured rather than derived bulk densities have been used to convert from SOC concentrations to SOC amounts, our observations question whether for permanent grassland in England, losses in SOC in recent decades reported elsewhere can be attributed to widespread environmental change.