HPLC is the best method to estimate the purity of natural compounds, reverse phase column is recommended for polar compounds.

Try TLC in different solvent systems , if you find a single spot it means the extracted  lutein is pure.If not then purify through coloumn or PTLC and check it again on TLC in different systems.  

Check if these publications are of any help

Just for purity check, it will be wise to use thin layer chromatography  (tlc) to save cost. Again use do tlc in at least two or three different solvents. If all solvents give one single spot, that is enough purity evidence. To quantify, you will need to use HPLC or GC if the extract is volatile. 

 I am sending you the paper kindly see the attached file

try looking up the published extinction co-efficient and then dissolve and dilute in the specified solvent a known amount of your solid. Then put in a cuvette and measure absorbance on a spectrophotometer at the specified wavelength. Then apply Beers Law. Solve for your unknown and calculate your purity. It will not be exact but should be close as marigold is mostly lutein with trace of beta carotene.

I have these values of absorbance for lutein :

- in hexane  420, 446, 474 nm

- in benzene 429, 456, 485 nm

- in dioxane 429, 453, 481 nm ( € 100000, 152000, 142000),

Remember to check your TLC at least in three solvent system or run a bi dimensional TLC.

Thanks to Olga we now can run the sample in dioxane at one of these wavelengths and calculate the concentration. I forgot a very important point, is that lutein is recoverable from saponified marigold oleoresin, because the marigold makes mostly lutein esters. If your biomass has not been saponified, you need to reflux in methanolic KOH, as I recall, to knock the esters off the lutein. Then use a non-miscible solvent, ie hexane, to extract the lutein away from the saponification solvent.

even without the saponification, the extinction coefficients above should give a good estimate of the lutein content to be found in the marigold lutein esters.