Bamboos are typical examples of highly synchronized semelparous species. Their mass-flowering events occur at supr-aannual intervals but they sometimes flower on a small scale in off-years. If some bamboo ramets (culms) of a genet flower and die in off-years, whereas other culms of the same genet do not flower synchronously, the genet can still survive blooming in an off-year and could participate in the next mass-flowering event. At genet level, the effect might be similar to that achieved by synchronously reproducing iteroparous plants. In addition, if multiple genets flower simultaneously in offyears, cross-pollination will be promoted. However, it is not known whether all the culms in a genet flower synchronously and whether multiple genets flower in off-years. We determined the clonal structure of three temperate dwarf bamboo species, i.e., Sasa senanensis, S. kurilensis, and S. palmata, at 24 off-year flowering sites and the surrounding areas in northern Japan using seven microsatellite markers. We also estimated seed set at seven of the sites and self-pollination rates at five sites to determine off-year reproductive success. Next, we investigated whether seed sets at the culm level were related to flowering area and/or number of flowering genets, using generalized linear mixed-effect models (GLMMs). Multiple genets flowered at 9/24 flowering sites. We found that 40/96 of the genets identified had some flowering culms. Non-flowering culms were present in 24/40 flowering genets. Seed set was in the range 2.2%-12.5% and the self-pollination rate was 96.3%. In the best GLMM, seed set increased with flowering area. Seeds were produced in off-years, but cross-pollination was rare in off-years. We suggest that some dwarf bamboos may exhibit iteroparity or imperfectly synchronized semelparity at the genet level, a characteristic similar to that of other reproductively synchronous plants. We also found synchronous flowering of a few genets even in off-years.