Hepatic granulomatous inflammation is one of the key pathological lesions of a patent Schistosoma mansoni infection. This study was concerned with the sequential induction, formation and eventual modulation of the schistosome egg granuloma in the mouse, which develops schistosome-induced hepatic fibrosis, and the olive baboon, which usually does not. Six baboons were each infected with 1500 S. mansoni cercariae and liver biopsies were collected at weeks 6, 9, 13 and 17 post-infection (p.i.). The mice (n = 25) were each infected with 100 cercariae and killed in groups of five at weeks 6, 9, 12, 18 and 21 p.i. Peak granuloma size was observed at week 6 p.i. in baboons (mean 355 ± 65.6 μm) but at week 12 p.i. in mice (299 ± 40.5 μm). Eosinophils were more abundant in the baboon (60.6 ± 8.9%) than in the mouse (41.2 ± 8.4%) at the time of maximal granuloma size (P<0.01). Neutrophils formed 21.1 ± 6.3% of peak mouse granulomata but were virtually absent in baboon granulomata. A feature of the modulating baboon granulomata was the emergence of multinucleated giant cells (MGCs); modulating mouse granulomata, on the other hand, were characterized by infiltration of fibroblasts and collagen deposition. Thus, by week 21 p.i. mouse granulomata were 92 ± 16.0 μm in diameter and well delineated by concentric layers of fibrous tissue. Granulomata, however, were present in only two of the baboons at week 17 p.i. (44 ± 61.2 μm in diameter). The other four had peri-portal cellular infiltration without granuloma formation, implying that baboon granulomata resolve spontaneously. These data suggest that high tissue eosinophilia and MGC formation are particularly efficient in bringing about the destruction of schistosome eggs and subsequent resolution of the egg granuloma without fibrosis. In conclusion, the baboon model more closely mimics the pathogenesis observed in man than does the mouse model. (C) 2000 Harcourt Publishers Ltd.