Male-male combat is a ritualized behaviour occurring mainly during the mating season (Greene, 1997; Schuett, 1997; Schuett et al., 2001) which is widespread among snake families including Boidae, Elapidae, Viperidae and Colubridae (Shine, 1978, 1994). Male-male combat is an example of intrasexual selection in which males compete for access to females. Typically, the two opponents attempt to physically subdue one another by exerting pressures (e.g., pushing, flipping, entwining) on the anterior part of their body in an attempt to bring their adversary to the ground. (Carpenter, 1977). Among boid snakes male-male combat is recorded in only one erycinae (Eryx colubrinus – see Schuett et al., 2001), 21 pythoninae (see Shine, 1994; Schuett et al., 2001) and 14 boinae species – nine of these being Neotropical (see Shine, 1994; Schuett et al., 2001). The genus Epicrates includes ten species and combat is recorded for six, including the Brazilian E. c. cenchria (Shine, 1994; Schuett & Schuett, 1995; Schuett et al., 2001). Despite the existence of many records in boid snakes, detailed descriptions of combat behaviour are only available for the pythoninae (Python molurus ; Baker et al., 1979) and the boinae Sanzinia madagascariensis. Here we provide a description of combat in three captive male Epicrates cenchria crassus. This mainly terrestrial species feeds on small mammals (Pizzatto L., unpubl. data), and inhabits the Brazilian cerrado, a savanna-like vegetation, where it is probably associated with riparian forests (Peters et al., 1986; Henderson et al., 1995). The three snakes came from unknown localities in São Paulo State, southeastern Brazil and were maintained in captivity in Bandeirantes city (-50°03'W, -23°02'S), Paraná State, southern Brazil. Male 1 and male 2 were obtained in 1999 and male 3 was obtained in early 2001. These snakes were maintained in individual cages (1.0 m x 0.50 m x 0.40 m) made of wood with screen in the front. Cages were contiguous and disposed side by side in the same room, thereby preventing visual contacts between the experimental subjects. Room temperature was ca. 30–32°C in spring/summer and 22–25°C in autumn/winter. We used heating pads when the temperature fell below 25°C, and sprayed the cages with water when the humidity was low. Water was provided ad libitum, and snakes were fed once a week (one or two young rats). We measured each snake in June 2004: Male 1 was 1130 mm in SVL and weighed 845 g, male 2 was 910 mm in SVL and 800 g, and male 3 was 1060 mm SVL and 910 g. All males were adults (L. Pizzatto, unpl. data). We used the terminology used by Carpenter et al. (1976) for the combat phases. In April/May in 2001, we observed high snake activity in the cages and decided to introduce males to one another, simultaneously, in the same cage. Then, male 1 and 3 started a combat behaviour. Although male 2 was in the cage with the other 2 snakes it did not engage in combat. The end of the combat was initiated when the loser (male 3) withdrew itself from the fight and became inactive. After trials, all snakes were returned to their individual cages. Male-male ritualized combat in the Brazilian rainbow boa, Epicrates cenchria crassus