A series of small ponds was constructed to determine whether dragonflies would oviposit in them. Four pond sizes - 0.3, 0.6, 0.9, and 1.2 m² - were replicated and randomized in each of six blocks. Ponds in three of the blocks were planted with cattail (Typha sp.) and three with spike rush (Junca sp.). Sampling revealed that of 273 nymphs collected, 267 were Pantala flavescens (F.) and six were Orthemis ferruginea (F.), although adults of 15 other dragonfly species were seen in the study area. Dragonfly nymphs were not found in any of the smallest size of pond. Significantly more nymphs were found in ponds planted with cattail than rush (215 vs 52), but the nymphs grew larger in ponds with rushes. The mean number of nymphs was larger in the 1.2-m² ponds (21.8) than in the 0.6-m² ponds (8.7), while 0.9 m² was intermediate (15.0). However, when the number of nymphs per unit area was compared, there were no differences between any of the pond sizes within a vegetation class -- cattail or rush. Between classes, more nymphs per unit area were in ponds planted with cattails. Because the one species of dominant dragonfly was not representative of distribution of dragonfly species in normal-sized ponds and probably was the only dragonfly found in any similar experiment, use of small reference ponds of these sizes was not feasible for assessing dragonfly diversity and abundance.