Population matrix models are used to assess population viability (Caswell, 2001). Given a fully parameterized population matrix model, managers may use these matrices to inventory the population, diagnose issues with population viability, investigate hypothetical management activities, or prescribe action items to stem or reverse population decline. But what if managers do not have the data to fully parameterize a matrix model? This work explores the use of a combinatorial optimization algorithm (Korte & Vygen, 2018) in backfilling the parameters of a matrix model given adult time series data (Ding et al. 2008). The method is explored using simulated adult trajectories of Northern Spotted Owl (Noon & Biles, 1990), with several noise types and variance stabilizing transformations (Dennis et al., 2001). The algorithm performance is compared against the performance of conditional least squares (Klimko & Nelson, 1978) and ordinary least squares (Fox, 2016), and all are compared against truth. The algorithm assumes a deterministic matrix, that the matrix elements remain static over the 29-year time periods, and that the time series data is free from sampling error.