Several biological parameters were measured in 31 heifers naturally infected with Fasciola hepatica during one grazing season in the Belgian Ardennes. A forecast model based on daily temperature used to assess the risk of fasciolosis was fitted to this assay. Cattle were turned out to two pastures. Each pasture was divided into two plots: one was treated with calcium cyanamide and the other was left untreated. The Lymnaea truncatula snails were counted on three different occasions. The results indicated a poor molluscicide efficiency. Body weight gains, anti-Fasciola antibody levels, faecal egg counts, levels of sorbitol dehydrogenase (SDH) and gamma-glutamyl transferase (gamma GT), packed cell volumes, white blood cells and differential leucocyte counts were determined monthly. No statistically significant difference was observed between animals from the two plots regardless of the recorded data. No correlation was found between body weight gains and other biological data. The sampling date had a significant effect on the antibody responses within a same group, and on the enzymatic levels for all groups combined. The forecast results were consistent with the recorded data. Temperature was a major bioclimatic constraint on the transmission of life cycle, and risk of infection occurred mainly in late spring (May/June) and in early September. Current results might be used to issue advice on the need for flukicide treatment of cattle. The indicators of the infection considered alone were useless and it is concluded that herd diagnosis of fasciolosis may rely on the rise of specific antibody levels, possibly associated with an increase in hepatic enzyme activities.