I tested the hypothesis that there is a correlation between inundation period and net primary production (NPP) in estuarine salt marshes through an observational experiment at South Slough National Estuarine Reserve, Oregon. I estimated primary production and inundation period for selected monotypic plant communities in (1) a naturally restored wetland that has restricted circulation with a dike system, and (2) a nearby control wetland with open circulation over the 1992 growing season. Carex lyngbyei (Caly) plots in Rhodes Marsh, the site with restricted circulation, were inundated for 1671 hours, and Distichlis spicata (Disp) plots were inundated for 1004 hours. In Stratigraphy Bay, the site with open circulation, the Caly plots were inundated for 1057 hours and the Disp plots were inundated for 714 hours. Restricted circulation Caly plots were inundated 58% longer than open circulation Caly plots. Restricted circulation Disp plots were inundated 41% longer than open circulation Disp plots. The drainage time for the two systems was constant regardless of the high-water elevation. Change in the inundation period with different high-water elevations was caused by differences in filling time and not draining time. The Caly community in Rhodes Marsh with restricted circulation had a net above ground production value of 659 g/m??/yr. The Disp community had a net above ground production value of 546 g/m??/yr. The Caly community in Stratigraphy Bay with open circulation had a net above ground production value of 1210 g/m??/yr. The Disp community had a net above ground production value of 1197 g/m??/yr. The Caly plots in the system with restricted circulation had 54.5% of the net primary production of the Caly plots in the open circulation system. The restricted circulation Disp plots had 45.6% of the net primary production of the open circulation Disp plots. My estimates of inundation periods and above ground net primary productivity of selected plant communities from a wetland site that has restricted circulation with a dike system, and from a site with open circulation, supported the hypothesis that inundation period and net primary productivity are negatively correlated (R?? = 0.972, slope = -0.90 for Caly fit to a linear regression model, n = 6, R?? = 0.980, slope = -2.24 for Disp fit to a linear regression model, n = 4). Thus, I confirmed the model-based assumption that increasing circulation yields an increase in net primary productivity.