The operationally defined fraction of soil organic matter (SOM), glomalin-related soil protein (GRSP), has been reported to be particularly stable. This is the first study to test this hypothesis. Archived soil has been used from the long-term bare fallow trial, known as the 42-Plots, in Versailles, France. With no input from plants for 80 years, changes in GRSP must result from either turnover of SOM or net mineralization. Intrinsically, stability of protein would lead to progressive enrichment in GRSP. Soils from treatments that influenced pH, soil C:N ratio and physical stability were studied. Soil having received manure amendment was also studied to give contrasting microbial activity. Both GRSP and soil organic carbon (SOC) contents declined over the 80-year period, with the exception of the manure treatment. GRSP and SOC contents were positively correlated. SOC was continuously enriched in GRSP in the control treatment, suggesting greater stability of GRSP. The SOC in manure-amended soil was also progressively enriched in GRSP, but this can be attributed to the GRSP content of added manure. For the other treatments, the ratio GRSP:SOC fell continuously (CaCO3 amendment) or fell to a minimum value after 45 years (KCl amendment) or levelled off after a progressive increase ((NH4)2SO4 amendment). Trends are compared with measured soil properties and changes in microbial activity inferred from soil pH. No reliable independent assessment of microbial composition of these archived soils is possible. The results cannot be explained by inherent chemical or physicochemical stability of GRSP. We conclude that recycling of SOM continually produces soil protein. The net change in GRSP content is a complex function of SOM cycling and protein mineralization. These findings add to the increasing body of evidence that the operationally defined fraction of SOM known as GRSP does not predominantly originate from arbuscular mycorrhizal fungal activity. Highlights: Processes determining SOM enrichment in GRSP in C-depleted soils are not understood. GRSP decreased along with SOC, but relative rates of decline varied with amendments. Levels of GRSP appear to be due to organic matter turnover as well as mineralization. GRSP does not appear to originate from arbuscular mycorrhizal fungal activity.