Abstract Background Testing for anti-Toxoplasma immunoglobulin (Ig)M is of main importance in the context of pregnancy to promptly alert to an acute maternal infection prior to the detection of IgG and to identify infected newborns. Their absence helps exclude a recent maternal infection in the presence of IgG. Methods The performance of a Toxo IgM immunocapture prototype assay (bioMérieux, France) was compared with that of the VIDAS^® Toxo IgM and the ARCHITECT^® Toxo IgM (Abbott, Germany) assays at Grenoble and Lyon (France). A total of 1446 sera were sampled from (i) 1054 pregnant women found by routine workup to have no infection (n = 843), an acute infection (<4 months) (n = 28) or a chronic infection (>4 months) with residual (n = 120) or no IgM (n = 62); (ii) 50 three-serum panels sampled immediately after a maternal seroconversion; (iii) 242 samples taken in 41 children with a congenital toxoplasmosis (n = 122) and in 40 uninfected children (n = 120). Results In pregnant women, the overall agreement with the VIDAS^® assay was 99.23% (CI: 99.16–99.27) and that with the ARCHITECT^® assay was 99.14% (CI: 99.07–99.17). Sensitivity of the Toxo IgM prototype assay was 100% (CI: 87.66–100.00) and specificity was 99.64% (98.96–99.93). In acute maternal infections, IgM assays were detected as early with the prototype as with the other two. In the congenitally infected children, IgM were detected on their first sample in 25/40 with the prototype vs. 23/40 with the VIDAS^® test. No uninfected child had positive IgM. Conclusion The prototype performed comparably to the ARCHITECT^® and VIDAS^® Toxo IgM assays for the diagnosis of maternal and congenital toxoplasmosis.