background: obstetrics injuries can be defined as fetal lesions suffered by the fetus during labor or expulsion. objective: identify risk factors for fetal birth trauma. method: case-control study conducted from july 2004 to december 2005. we analyzed maternal and fetal variables using software spss 19.0. the variables showing a significance level <0.10 in the bivariate analysis were included in the logistic regression analysis. we used the block entry model for selection of the final regression model. results: during these period, we identified 2137 births and 26 (1.2%) were related to fetal birth injury. clavicle fracture was the most frequent injury (n = 14; 53.8%), followed by cephalohematoma (n = 5; 19.2%). umbilical cord blood gas analysis did not show statistical significance. in the regression analysis of the variables that had significance level <0.10, the only independent variables associated with fetal birth injury were vaginal delivery (or-a: 11.08; 95%ci: 2.45-49.98; p=0.002) and thorax circumference > 33 cm (or-a: 3.36; 95%ci: 1.35-9.73; p=0.010). conclusion: the risk factors for fetal birth injury were vaginal delivery and chest circumference equal to or larger than 33 cm. other variables were not associated with the outcome in discussion.