Canola was studied as an alternative source to traditional microbial systems as a recombinant protein production host. Various chromatographic methods were used in this work to systematically characterize the native canola protein elution profiles, and different genetically engineered proteins were selected to explore the opportunities presented for effective protein recovery;Native canola protein was eluted into two major peaks in linear salt gradient elution on cation-exchange chromatography. T4 lysozyme and its mutants, by both point mutation and fusion, were used as model proteins to investigate the charge effect on protein elution. It was found that the single mutant T4 lysozyme was eluted in the low canola background valley between the two major canola protein peaks under the experiment conditions, and the purity of collected lysozyme was more than 90%;Canola protein elution profile presented three possible target sites for selective recombinant protein purification on a strong anion exchanger. beta-Glucuronidase (GUS) and fusions were used as the model proteins. Wild-type GUS was found to be eluted roughly at the first target site with the least eluent salt concentration, and GUS fusions with more than 10 aspartates in the tail were moved to the second site (higher salt concentration) with lower canola protein background. The third site with the highest salt elution was out of reach. The enrichment factor of fusion GUS at the second site was three to four times higher than that of wild-type GUS;Stoichiometric displacement model was used to characterize the series of GUS proteins (wild-type GUS and its fusions) under isocratic elution. It was found that the fusion with 15 aspartates did not follow the trend of change in protein specific parameters, Z and I, as the number of aspartates was less than 10. The Z and I values were used in an equation and lumped dispersion model to predict the protein elution under various gradient slopes. Both methods could reasonably well predict the protein retention and the influence of gradient change on the protein elution, and the simulation could also pick up the shape of the protein peaks;Immobilized metal affinity chromatography (IMAC) was also used to explore the possibility of purifying proteins with poly-histidine fusions. GUS-(his) 6, as our model protein, could be purified to almost homogeneous purity on Co2+ columns with iminodiacetate (IDA) and nitrilotriacetate (NTA) as the immobilization ligand. The recognition of metal ions on protein surface histidine distribution was found following the order Cu2+ > Ni2+ > Zn2+ > Co2+. The binding mechanism was proposed to describe the interaction between ploy-his tagged protein with immobilized metal ions when IDA and NTA were used as the chelating ligands.