An enclosure experiment in the shallow, subtropical Lake Donghu, China, was performed in the summer of 2001 to examine the effect of TN:TP (total phosphorus) ratios and P-reduction on the occurrence of Microcysitis blooms. The treatments were performed with enough amounts of N but with different amounts of P in the water column and sediment. Microcystis blooms occurred in the enclosures either with an initial TN:TP <29 or TN:TP>29 where the nutrients (N, P) were high enough. Microcysitis blooms never occurred in the treatments with low P concentration in spite of the presence of sufficient N. The P-rich sediments served as an important source for the P supply in the water column, and such a process was activated greatly by the outburst of Microcystis blooms which pumped up selectively P from the sediments and thus decreased the TN:TP ratios. Therefore, the low TN:TP ratio is not a cause but rather a result of Microcystis blooms.