Pelobionts (Archamoebae sensu Cavalier-Smith 1991) lack mitochondria and are mostly free-living, heterotrophic, amoeboid, flagellated protists that inhabit organically enriched, microoxic, or anoxic freshwater and marine environments (Schulze 1877;Penard 1921;Brugerolle 1993). Because of their ultrastructural simplicity and the basal position of many amitochondriate parasitic protists in ribosomal RNA phylogenies (Sogin 1997), pelobionts were hypothesized to represent one of the earliest diverging eukaryotic lineages (Margulis 1970;Whatley 1976;Griffin 1979;Cavalier-Smith 1983;Griffin 1988;Whatley and Chapman-Andresen 1990;Cavalier-Smith 1991;Brugerolle 1993;Patterson 1994). However, molecular data support conflicting hypotheses about the monophyly and phylogenetic placement of pelobionts. Early analyses of small subunit rRNAs (SSU rRNA) (Hinkle et al. 1994) and partial sequences of large subunit rRNAs (LSU rRNA) (Morin and Mignot 1995) show that the pelobiont Mastigamoeba balamuthi (ϭPhreatamoeba balamuthi [Simpson et al. 1997]) diverged relatively late among eukaryotes. In contrast, a phylogenetic analysis of DNA-dependent RNA polymerase II (RPB1) genes (Stiller, Duffield, and Hall 1998) placed a sequence from an organism identified as M. invertens as one of the most basal eukaryotes.Discrepancies between inferences from cytology and RPB1 (Stiller, Duffield, and Hall 1998) versus LSU rRNA (Morin and Mignot 1995) and SSU rRNA data (Hinkle et al. 1994) prompted Stiller andHall (1999) to examine the position of M. invertens in SSU rRNA trees. Their initial maximum likelihood (ML)-based analysis did not cluster M. invertens with the other pelobiont sequences, nor were they basal to other eukaryotes. Stiller and Hall hypothesized that long-branch attractions (LBA) or skewed base compositions in rRNAs account for conflicting phylogenies of SSU rRNAs and RNA polymerases. When the rapidly evolving lineages were removed from that analysis, the two pelobionts formed a monophyletic group basal to other eukaryotes, but the GC content of half of the basal branches was not aberrant and was unlikely to impact the analysis.To reexamine the evolutionary position and monophyly of pelobionts, we obtained new pelobiont sequences that were PCR amplified (Sogin 1990) We performed the likelihood analysis shown in figure 1 using PAUP* version 4.0.0b4a (Swofford 1999) with TBR branch swapping and 117 random-addition replicates. Modeltest version 3.0 (Posada and Crandall 1998) revealed that all models more simple than the general time-reversible model of nucleotide substitution, with invariable sites and a gamma distribution of among-site rate variation (GTR ϩ I ϩ ⌫) had a significantly poorer fit to the data. This model was used for both ML and minimum evolution (ME) distance bootstrap analyses with 100 and 1,000 replicates, respectively. Empirical base frequencies were used, the proportion of invariable sites was estimated at 0.173, and the value of the shape parameter for the gamma distribution was 0.615. Under ML we discovered 10 ...