A mathematical approach was used to model seasonal variation of cyathostomin faecal egg output, and in simulating anthelmintic control strategy. The model is based on parameters of biological development of cyathostomins and climatic data. A good fit of the model prediction to the field data was obtained after some parameter adjustments. The development rate of ingested larvae to egg laying adults, survival time of adults and the assumption made in modelling the peak pasture larval availability were the main driving forces for the model prediction to fit to the observed data. The apparent fit of the model prediction to the field data obtained after parameter adjustment generally indicated some major differences between donkeys and horse in their reaction to the parasite and/or between cyathostomins of donkeys and horses. The results of the stimulation of the effect of various protocols for the timing and frequency of anthelmintic treatment on the adult cyathostomin worm burden have shown that treating donkeys only once in a year or a combination of once in a year followed by every two or even four years can substantially reduces and maintains the parasite burden far below the pre-treatment level for many years. Generally the study made has revealed that the non-strongyle gastrointestinal parasites of donkeys are highly prevalent and have a high pathogenic potential, and the findings of cestodes and trematodes are not accidental or unusual, as previously suggested.