Thalassophryne nattereri fish envenomation is commonly reported among fishermen and bathers in the Brazilian north and north-east coast, estimated at hundreds of accidents per year (Haddad et al. 2003). On Ceara State seaside, 16 cases of envenomation by T. nattereri were notified, from 1992 to 2002 (Faco et al. 2005). According to Fonseca and Lopes-Ferreira (2000), the palm of the hands or soles of the feet are the areas most commonly affected in the victims when pierced by spines connected to the fish's venom glands. Envenoming symptoms are readily evident, including local oedema, erythema and severe pain followed by intense necrosis and a markedly inefficient healing response. This problem of the inefficient healing is very important for the evolution and treatment of the accident, which is devoid of specific drug treatment (Lopes-Ferreira et al., 2000; Haddad et al. 2003). According to this, the symptoms advance, taking weeks or even months for complete recovery (Fonseca & Lopes-Ferreira 2000). Investigations carried out in our laboratory have shown that the dramatic symptoms of T. nattereri envenomation are related to kallikrein–kinin cascade, identified as the major mechanism involved in the nociception as well as the oedematous response to venom, as only the administration of novel tissue kallikrein inhibitor (phenylacetyl-Phe-Ser-Arg-N-(2,4-dinitrophenyl)-ethylenediamine – TKI), and not PKSI-527 and SBTI, specific and non-specific plasma kallikrein inhibitors, respectively, reduced these clinical manifestations (Lopes-Ferreira et al. 2004). Thalassophryne nattereri induces direct damage to skeletal muscle plasma membrane together with thrombosis and other alterations in the microvasculature of mice, without inducing haemorrhage (Lopes-Ferreira et al. 1998 2001). Analyses of the cremaster muscle showed that the venom elicits a peculiar alteration of the microcirculation with intense vascular congestion, thrombosis in venules and focal transient constrictions in arterioles (Lopes-Ferreira et al. 2002). In contrast to other models of myonecrosis, in which the microvasculature is not affected, phagocytosis of necrotic material is well advanced by 24 h and no remnants of necrotic muscle cells are observed after 1 week (Harris et al., 1975; Gutierrez et al., 1991; Morini et al., 1998. Curiously, low numbers of phagocytic cells during the first 24 h after)T. nattereri injection, and the presence of necrotic material which had not been cleared out 7 days after envenomation was described (Lopes-Ferreira et al. 2001; Lima et al. 2003). Nevertheless, the reason that the venom toxins favour delayed local inflammatory response is ill defined. This study was carried out to describe leucocyte movement into mice tissue after the injection of T. nattereri venom in order to investigate the role of inflammatory mediators, metalloproteinases and cell adhesion to extracellular matrix (ECM) components present in this envenomation.