In a survey of marine annelids for myxosporean infection in Charleston Harbor, South Carolina, we collected 3,214 polychaetes from 21 families and found infections in 6 spionid individuals. Based on gross morphology and COI sequencing, all infected spionids were identified as Streblospio benedicti. Infection prevalence was 0.8% (6/734) of that species of spionid, and 0.2% of all 3,214 polychaetes examined. Pansporocysts contained 8 actinospores and developed in the tegument of the annelid host. This is the first myxozoan infection recorded from this polychaete species, second in the family, and the first marine myxozoan found in the Americas. It is the first marine species found to develop in the tegument of its annelid host; a site of development observed only once before, in Ceratonova shasta infections of its freshwater sabellid polychaete host. Mature actinospores were morphologically simple, truncated ellipsoids, lacking processes or ornamentation, 9.0 ± 0.5 µm × 6.0 ± 0.4 µm. Their sack-like shape was similar to 9 of the 12 actinospores described previously from polychaetes; 10/12 had been and ascribed originally to the morphological collective group Tetractinomyxon despite 9 of these having few similarities to the original description of this group. We propose to name the simple, spherical to ellipsoidal spore morphotype Saccimyxon to encompass both our novel actinospore and the 9 other sack-like polychaete actinospore types in the literature. In the present study, 18S rDNA sequencing demonstrated that the myxozoans that infected the 6 spionids were genetically the same species (type sample 1,737 nucleotides, GenBank accession number MH791159) and was not >95% similar to any sequence in GenBank. Phylogenetic analysis showed that the myxozoan species we encountered is basal to the kudoids and thus likely to have a morphologically simple myxospore stage with fewer than 4 valves. However, without a genetic match, the presumptive vertebrate host remains unknown.