This study was aimed to set up and evaluate a quantitative method for detecting lumbrokinase level in plasma. The lumbrokinase was used to immunize rabbit and BALB/c mouse for preparation of rabbit or mouse-derived polyclonal antibodies, and then the standard curves were drawn up by detecting the lumbrokinase diluted in PBS using the double antibody sandwich ELISA. This method further was analyzed for its specificity, precision and recovery rate. This established double antibody sandwich ELISA was used to assay the lumbrokinase in human plasma, and the assayed results were assessed. The results showed that a double antibody sandwich ELISA for the detection of lumbrokinase has been established. And the standard curve fitting R value > 0.99, the precision assessment showed that the measured values of coefficient of variation (CV) in 3 batches were all < 15%; recovery assessment in 3 batches showed that all the measured recovery rates were > 80%; the quantitative low limit was assessed as 5 ng/ml (precision CV < 15%, recovery rate > 85%). It is concluded that this method is consistent with the criteria stipulated by the Pharmacopeia, which provides a reliable measurement method for quantitative detection of plasma lumbrokinase in clinical trials.