Abstract:The accuracy of artificial nests in representing natural patterns of nest predation has been widely studied in temperate regions and egg size is one of the most tested sources of bias. In the neotropics, experiments with artificial nests usually used larger than natural eggs, despite suggestions in literature that the eggs should be similar to those of the local species. Here, we tested the hypothesis of spatial-temporal variation in predation risk of artificial nests in relation to egg size. We used quail (Coturnix coturnix; large), Chestnut-bellied Seedfinch (Oryzoborus angolensis; small) and plasticine (both sizes) eggs placed in artificial nests. We analyzed daily nest survival using models in the Program MARK. The best-fit model included the effects of egg size and the reproductive period on daily survival of artificial nests. Nests with large eggs had greater Daily Survival Estimates (DSE) than nests with small eggs during two times and DSE were smaller early in the reproductive period. DSE of 0.82 (95% CI = 0.76 to 0.86) and 0.91 (0.87 to 0.93) were recorded for large eggs at the beginning and end periods compared to 0.72 (0.65 to 0.78) and 0.84 (0.79 to 0.88) for small eggs. Birds were more important predators (40% of records) than mammals and reptiles based on signs left on the plasticine eggs. Our results support the hypothesis that larger eggs are less depredated than small ones in artificial bird nests, even when main predators (birds) are different from those commonly observed in the Neotropics (mammals and reptiles). However, controlling the egg size in experiments with artificial nests may not be sufficient to obtain similar time patterns observed in natural nests studied in the Cerrado. Resumo: A acurácia de ninhos artificiais em representar padrões naturais de predação de ninhos foi amplamente estudada em regiões temperadas, sendo o tamanho do ovo uma das fontes de erro mais testadas. Nos Neotrópicos, os experimentos com ninhos artificiais geralmente usaram ovos maiores que os naturais, apesar da literatura indicar a necessidade do uso de ovos semelhantes aos das espécies locais. Aqui, testamos hipóteses sobre variação espaço-temporal no risco de predação de ninhos artificiais em função do tamanho do ovo. Para isso, usamos ninhos artificiais com ovos de Codorna (Coturnix coturnix; grandes), Curió (Oryzoborus angolensis; pequenos) e plasticina (ambos tamanhos). Analisamos a sobrevivência diária dos ninhos através de modelos gerados no Programa MARK. O modelo melhor ajustado aos dados incluiu o efeito do tamanho do ovo e do período reprodutivo na sobrevivência diária dos ninhos artificiais. Ninhos com ovos grandes tiveram maiores Estimativas de Sobrevivência Diária (ESD) durante os dois períodos, além disso, as ESD foram menores no início do período reprodutivo. ESD de 0,82 (95% IC = 0,76 a 0,86) e 0,91 (0,87 a 0,93) foram registradas para ovos grandes no início e fim da estação reprodutiva, comparado com 0,72 (0,65 a 0,78) e 0,84 (0,79 a 0,88) registradas para ovos pequenos. As aves fora...