It is critical to develop a cost-effective quantitative/semiquantitative assay for rapid diagnosis and on-site detection of toxic or harmful substances. Here, a naked-eye based semiquantitative immunochromatographic strip (NSI-strip) was developed, on which three test lines (TLs, TL-I, TL-II and TL-III) were dispensed on a nitrocellulose membrane to form the test zone. Similar as the traditional strip assay for small molecule, the NSI-strip assay was also based on the competitive theory, difference was that the analyte competed three times with the capture reagent for the limited number of antibody binding sites. After the assay, the number of TLs developed in the test zone was inversely proportional to the analyte concentration, thus analyte content levels could be determined by observing the appeared number of TLs. Taking aflatoxin B(1) as the model analyte, visual detection limit of the NSI-strip was 0.06 ng mL(-1) and threshold concentrations for TL-I-III were 0.125, 0.5, and 2.0 ng mL(-1), respectively. Therefore, according to the appeared number of TLs, the following concentration ranges would be detectable by visual examination: 0-0.06 ng mL(-1) (negative samples), and 0.06-0.125 ng mL(-1), 0.125-0.5 ng mL(-1), 0.5-2.0 ng mL(-1) and >2.0 ng mL(-1) (positive samples). That was to say, compared to traditional strips the NSI-strip could offer more parameter information of the target analyte content. In this way, the NSI-strip improved the qualitative presence/absence detection of traditional strips by measuring the content (range) of target analytes semiquantitatively.