In this paper we study the presence of calendar anomalies in the main Latin-American stock markets, for the 1993 to 2007 period. The literature has shown that the detection of those effects may depend on error distribution assumptions (Baker et al., 2008), and that their existence could be due to a problem of data snooping (Sullivan et al., 2001). In response to these problems, Cho et al. (2007) introduced a robust non-parametric test, which was adopted in this paper. Results show that calendar anomalies: Monday and Weekend Effects are present in the main Latin-American stock markets, and they are statistically significant.