The North American Marsh Bird Monitoring Protocol (NAMBMP) was implemented to monitor population trends for secretive marsh birds. Single season abundance and occupancy estimates from NAMBMP data assume closed populations of resident marsh birds; however, the assumption that migrant individuals are either not present or not detected has not been tested. Furthermore, limited information is available about the effects of other environmental or biological variables on vocalization rates of secretive marsh birds. The residency status of 70 Virginia Rails (Rallus limicola) and 25 Soras (Porzana carolina) was determined based on length of stay and time of year present and differences in call-response rates and call-types between resident and migrant rails were tested. No differences between call-response rates (49.3% = residents, 44.4% = migrants, x²₁ = 0.37, P-value = 0.55) or call-types (x²₂ = 3.48, P-value = 0.18) were found between resident and migrant Virginia Rails. Sample size of resident Soras (n = 2) was insufficient to test for differences between residents and migrants and neither resident Sora responded; however, migrant Soras were found to respond during call-response surveys (34.3%). Few environmental or biological variables influenced vocalization rates in this study; however, Virginia rails were more likely to vocalize when other conspecifics vocalized (β _i = 0.76, P-value = 0.030). This research suggests NAMBMP surveys in Ohio detect both resident and migrant Virginia Rails and Soras, and thus the population closure assumption is not met. Research testing the assumption of population closure for all marsh bird species at larger geographic scales is recommended and caution should be used when estimating single season occupancy or abundance using NAMBMP data.