SUMMARY: Although systematic treatments of Hydrozoa have been notoriously difficult, a great deal of useful information on morphologies and life histories has steadily accumulated. From the assimilation of this information, numerous hypotheses of the phylogenetic relationships of the major groups of Hydrozoa have been offered. Here I evaluate these hypotheses using the complete sequence of the 18S gene for 35 hydrozoan species. New 18S sequences for 31 hydrozoans, 6 scyphozoans, one cubozoan, and one anthozoan are reported. Parsimony analyses of two datasets that include the new 18S sequences are used to assess the relative strengths and weaknesses of a list of phylogenetic hypotheses that deal with Hydrozoa. Alternative measures of tree optimality, minimum evolution and maximum likelihood, are used to evaluate the reliability of the parsimony analyses. Hydrozoa appears to be composed of two clades, herein called Trachylina and Hydroidolina. Trachylina consists of Limnomedusae, Narcomedusae, and Trachymedusae. Narcomedusae is not likely to be the basal group of Trachylina, but is instead derived directly from within Trachymedusae. This implies the secondary gain of a polyp stage. Hydroidolina consists of Capitata, Filifera, Hydridae, Leptomedusae, and Siphonophora. "Anthomedusae" may not form a monophyletic grouping. However, the relationships among the hydroidolinan groups are difficult to resolve with the present set of data. Finally, the monophyly of Hydrozoa is strongly supported.