Open population models using capture-mark-recapture (CMR) data have a wide range of uses in ecological and evolutionary contexts, including modeling of stopover duration by migratory passerines. In using CMR approaches in novel contexts there is a need to determine the conditions under which open population models may be employed effectively. Our goal was to determine whether there was a simple a priori mechanism of determining the conditions under which CMR models could be used effectively in the study of avian stopover ecology. Using banding data (n = 188 capture histories), we examined the challenges of using CMR-based models due to parameter inestimability, adequacy of descriptive power (Goodness-of-Fit, GOF), and parameter uncertainty. These issues become more apparent in studies with limited observations in a capture history, as is often the case in studies of avian stopover duration. Limited sample size and sampling intensity require an approach to reducing the number of fitted parameters in the model. Parameter estimability posed the greatest restriction on the utility of open population models, with high parameter uncertainty posing a lesser challenge. Results from our study also indicate the need for >10 observations per estimated parameter (approximately 3 birds captured or recaptured per day) to provide a reasonable chance of successfully estimating all model parameters.