Objective: To reveal variation regular patterns in odor of unprocessed and processed Herba Siegesbeckiae. Method: Evaporability composition of unprocessed and processed Herba Siegesbeckiae was sampled from headspace of emanating and then response values were obtained.Discriminant factor analysis(DFA) and soft independent modeling of class analogy(SIMCA) were used to analyze characteristic parameters. Result: Electronic nose detection methods of unprocessed and processed Herba Siegesbeckiae were established

synthesis dry air supplied by air generators as carrier gas

flow rate 150 mL·min-1

headspace temperature 60 ℃

time 30 min

agitation speed 250 r·min-1

injection volume 2 mL

acquisition time 2 min

lag time 6 min.Discrepancy rate of response values from sensor LY2/gCT in odor of unprocessed and processed Herba Siegesbeckiae was -82%;Odor of unprocessed and processed Herba Siegesbeckiae could be divided into two areas by DFA discriminant model

the left area was unprocessed products

the right area was processed products;SIMCA was effective model

score was 92 by cross-validation. Conclusion: Odor of unprocessed and processed Herba Siegesbeckiae had a significant difference

and the difference could be digitalized according to odor characteristic parameters tested with electronic nose.