We tested whether a trade-off exists between tolerance to flooding and tolerance to drought in wetland plants by assessing biomass accumulation, relative growth rate (RGR), survival rate, and physiological response of three wetland plants growing in drought or flooded environments. In wetlands of China's Sanjiang Plain, Carex lasiocarpa typically occurs at low elevations (10–50 cm water depths), Carex limosa at medial elevation (10–30 cm depths), and Deyeuxia angustifolia at high elevation (0–10 cm depths). Plants of three species were subjected to flooding and drought treatments (25 days) in a greenhouse experiment. In the flooding treatments, biomass accumulation (range 0.007–0.031 g per plant) and survival rate (11%) were lowest in D. angustifolia. Relative growth rate (RGR) was highest in C. lasiocarpa (−0.006 d⁻¹), intermediate in C. limosa (−0.051 d⁻¹), and lowest in D. angustifolia (−0.118 d⁻¹) at the end of the flooding experiment. Alcohol dehydrogenase (ADH) activity in C. lasiocarpa and C. limosa increased with flooding time, whereas ADH in D. angustifolia did not vary over the experimental period. These results indicated that tolerance to flooding from highest to lowest among the three species was: C. lasiocarpa > C. limosa > D. angustifolia. In the drought experiment, RGR was lower in C. lasiocarpa, but higher in C. limosa and D. angustifolia. At this experiment's end, only D. angustifolia plants still survived. Under drought conditions, production of malondialdehyde (MDA, an indicator for assessing a plant's ability to tolerate drought) showed the same pattern as ADH production under flooded conditions for all species. These results indicated that tolerance to drought from highest to lowest among the three species was: D. angustifolia > C. limosa > C. lasiocarpa. Our experiments indicate that a trade-off exists between tolerance to flooding and tolerance to drought in the three marsh plants.