A population of the American crocodile (Crocodylus acutus) protected through local participation was studied at La Ventanilla estuary in the coast of Oaxaca, during 8 months (October 2003-May 2004). Size, structure and sex ratio of this population were evaluated in order to provide demographic information of the American crocodile using two capture-recapture methods. A total of 21 adults (8 females, 13 males), 11 subadults (5 females, 6 males), 88 juveniles (14 females, 62 males) and 23 neonates were captured, measured and individually marked. Goodness-of-fit tests for the Jolly-Seber model showed that our population data violated the assumption of equal probability of capture and resulted in a skewed estimate of population size. However, goodness-of- fit tests for the geometric estimator of the capture frequency model applied to the same capture-recapture data showed recapture frequencies conform to it. Population size estimates for this model were 29.6 ± 9.1 adults, 37.1 ± 6.6 subadults, and 682.5 ± 39.2 juveniles. Population structure does not fit the normal distribution (d=39.5, P>0.05), most individuals belonging to lower size classes (16.2% in class I, 61.5% in class II), with few subadults (9.1% in class III) and adults (13.3 %, I class IV). Overall sex ratio was significantly skewed towards males (3 males: 1 female). This analysis provides information for future management strategies for the C. acutus population at La Ventanilla estuary. [ABSTRACT FROM AUTHOR]