Fecal culture remains the standard for diagnosing Mycobaterium avium paratuberculosis (MAP) in individual cattle, however it generally requires 8-16 weeks to obtain results. The enzyme-linked immunosorbent assay (ELISA) has rapid turn-around time, although its sensitivity is poor, especially in subclinically infected cattle. It is generally accepted that as infection progresses mean ELISA response and sensitivity increases. Cattle infected with MAP have a long prepatent period in which no shedding of the organism occurs followed by intermittent then continuous shedding increasing in volume as disease progresses. While any cow infected with MAP is undesirable, only cattle shedding the organism pose an immediate threat to other cattle. These cattle generally have higher ELISA adjusted optical density (OD) values compared to cattle not shedding MAP. The objective of this study was to compare mean ELISA adjusted OD values of cattle with negative fecal cultures to cattle classified as either low (≤10 cfus) or high (>10 cfus) shedders and determine the positive likelihood ratios for each. Fecal culture results and ELISA adjusted OD values were evaluated from 2,578 adult cattle from six Michigan dairy herds over two years. Prevalence of MAP in these herds based on fecal culture ranged from zero to 42% with an average of 9.8%. The mean ELISA adjusted OD value for fecal culture negative cattle (n=2,362) was 0.023 (95% CI: 0.02-0.03). That for cattle classified as low shedders (n=158) and high shedders (n=58) was 0.201 (95% CI: 0.12-0.28) and 0.784 (95% CI: 0.52-1.05), respectively. Using the mean as the ELISA cutoff for each group, likelihood ratios for a positive test for fecal culture negative, low, and high shedders were 3.2, 9.7, and 1600, respectively. Given the ELISA adjusted OD, the likelihood ratios can be used to predict the probability of a cow shedding MAP, provided herd prevalence information is available.