The Toxoplasma gondii major surface antigen, called SAG1 or p30, is a highly immunogenic protein which has generated great interest as a diagnostic reagent, as a potential subunit vaccine, and for its role in invasion. Unfortunately, bacterial recombinant protein is grossly misfolded so that, for example, it is not effectively recognized by antibodies to native SAG1. To overcome this, we have turned to expression in CHO cells, using cotransfection of the SAG1 gene and the mouse dihydrofolate reductase (DHFR) gene into CHO cells that are DHFR-. SAG1 expression was amplified by methotrexate coselection of CHO cells in combination with fluorescence-activated cell sorting for SAG1 expression. The resulting population expressed recombinant SAG1 that is recognized by antiserum specific for natural, nonreduced SAG1, indicating that, unlike in bacteria, expression in CHO cells results in proper folding. Processing was at least partially correct in that, like natural SAG1, recombinant SAG1 was attached to the plasma membrane via a glycolipid anchor, although tunicamycin treatment was necessary to prevent N-glycosylation (SAG1 is not glycosylated in the parasite but does have a consensus N-linked site). Finally, purified recombinant SAG1 was recognized by human sera known to be reactive to toxoplasma proteins, indicating that this material has potential as a diagnostic reagent and possibly as a component of a subunit vaccine.