Questions We asked whether the competition among community components (a) destabilizes individual components; (b) increases the asynchrony of their fluctuations; and (c) stabilizes the total community biomass. Location Seminatural meadow in South Bohemia, Czech Republic. Methods We used biomass fluctuation data from a 13-year removal experiment. The plots used for this study contained, following experimental removal of other species, either mycorrhizal grasses only, mycorrhizal forbs only, or their mixture (grass "monocultures", forb "monocultures", and mixtures, respectively). Yearly peak aboveground biomass was available from ten blocks; biomass of the mixture plots was sorted into forbs and grasses. Temporal variability was characterized by coefficient of variation (CV) and synchrony by correlation coefficient, both calculated from the time series data. Results The variability of grass monocultures was higher than the variability of forb monocultures, which was slightly lower than the variability of mixtures. The variability of both grasses and forbs was higher in the mixture, where they are in competition with the other group, than in each of their respective monocultures. The correlation coefficients between the biomass of grass and forb monocultures within blocks were mostly positive, indicating that both groups tend to have similar physiological responses to weather fluctuations. The average correlation coefficient between the forbs and grasses in the mixture plots was significantly negative, thus reflecting the effect of competition between these two community components. The CV of the sum of forb and grass monocultures was similar to the CV of mixture plots. Conclusions Only competition between grasses and forbs leads to their negative mutual correlation, i.e., to compensatory dynamics. Our results support the hypothesis that competition has a destabilizing effect on individual community components (functional groups in our case) and increases asynchrony of their fluctuations. We have not found its stabilizing effect on the total biomass.