Whole-saliva IgA appears like an attractive noninvasive readout for intestinal immune induction after enteric infection or vaccination, but has failed to show consistent correlation with established invasive markers and IgA in feces or intestinal lavage. For reference, we measured antibodies in samples from 30 healthy volunteers who were orally infected with wild-type enterotoxigenic Escherichia coli. The response against these bacteria in serum, lavage, and lymphocyte supernatants (antibody-in-lymphocyte-supernatant, ALS) was compared with that in targeted parotid and sublingual/submandibular secretions. Strong correlation occurred between IgA antibody levels against the challenge bacteria in sublingual/submandibular secretions and in lavage (r=0.69, P<0.0001) and ALS (r=0.70, P<0.0001). In sublingual/submandibular secretions, 93% responded with more than a twofold increase in IgA antibodies against the challenge strain, whereas the corresponding response in parotid secretions was only 67% (P=0.039). With >twofold ALS as a reference, the sensitivity of a >twofold response for IgA in sublingual/submandibular secretion was 96%, whereas it was only 67% in the parotid fluid. To exclude that flow rate variations influenced the results, we used albumin as a marker. Our data suggested that IgA in sublingual/submandibular secretions, rather than whole saliva with its variable content of parotid fluid, is a preferential noninvasive proxy for intestinal immune induction.