Entamoeba histolytica isolates have been shown to fall into two groups based on isoenzyme analysis. These groupings ("pathogenic" and "nonpathogenic") correlate well with the clinical course of the infection. A controversy exists over whether isoenzyme patterns are stable or whether under certain circumstances an isolate can convert from one form to the other. Resolution of this uncertainty is of importance since the nonpathogenic pattern has never been observed in amebae isolated from cases of active disease. This implies that, if the patterns are stable, carriers of amebae with this nonpathogenic pattern may never develop invasive disease. Although we set out to study isoenzyme conversion, we have been unable to replicate the two published accounts of this phenomenon. We have examined all of the variables proposed to be involved in the triggering of conversion, both individually and in combination. In none of the experiments was an alteration in the isoenzyme pattern observed. We now believe that isoenzyme patterns are stable and that all available evidence, other than the reported conversions, points to pathogenic and nonpathogenic E. histolytica being distinct species. © 1992.