Estimation of life history parameters is an important part of evolutionary ecology. However, estimation of mortality rates can take decades in long-lived species. Therefore, methods for estimating mortality rates from animal remains are desirable because they enable the researcher to investigate mortality rates over a much shorter period of time. In this study, I investigate whether it is possible to accurately estimate mortality from a population of guanaco (Lama guanicoe) skulls in the absence of any other data. I also investigate the problem of missing data from skulls since weather, predation, scavenging, and fire may damage the skulls before collection. Guanaco skulls collected from 1978 to 1998 in Torres del Paine National Park, Chile were examined and classified as adult (all permanent teeth fully erupted) or juvenile (2 years of age or less). Tooth eruption patterns and 9 measurements were collected from the juvenile skulls. Those juvenile skulls which had sufficient teeth were aged by tooth eruption and examined for correlations between skull size, shape, and age. Missing data values were estimated through imputation and it was found that no skew was introduced into the data if skulls which had at least five of the eight measurements correlated with age present. Allometry was found to be subtle and isometric growth is therefore hypothesized. From skull size, it was possible to estimate age by linear regression on a measure of size (PC1). Age was estimated for those skulls for which tooth eruption data was incomplete or unavailable and which had at least five of the eight age-correlated measurements available after the missing values were imputed. From the 278 aged juvenile skulls and 392 adult skulls in the collection, mortality rates were calculated by life table analysis and compared to those found in two previously published studies. It was found that this study and the previously published skull-based study underestimated juvenile mortality when compared to the mortality rate estimated from the published radio-collar study. However, missing skull measurements can be imputed and size and age can be estimated from skull size if a reference population is available.