Desiccation of Tortula ruralis was achieved rapidly by placing the moss on the laboratory bench, or more slowly by placing it in desiccators with atmospheres of high relative humidities. Unlike the rapidly desiccated moss, the slower desiccated moss retained no polyribosomes in the dehydrated state, although polyribosome reformation and protein synthesis resumed on reintroduction of the moss to water. Protein synthesis commenced on rehydration of the slower desiccated moss at a greater rate than on rehydration of the faster desiccated moss. A lack of correlation between endogenous ribonuclease activity and polyribosome levels extracted from the moss suggests that the observed reduction in polyribosomes during desiccation was not due to their degradation but was more likely a consequence of stress-induced restriction on reinitiation of existing messenger RNA. The observed protein synthesis on rehydration of the moss was largely independent of any prior RNA synthesis.