Estimating of the molar exctinction coefficient (I assume at 280 nm) and the concentration requires the knowledge of your protein (the best would be the amino acid composition). Then you can use some online tool to estimate the parameters (link bellow).

http://web.expasy.org/protparam/

Here is a paper on the subject - http://onlinelibrary.wiley.com/doi/10.1002/pro.5560041120/pdf

I have used the classical method for a single subunit protein. For that, you would prepare a solution of the purified protein. Run an absorption spectrum and record the absorbance at 280 nm. Using the same solution or a carefully diluted solution of it (keeping track of the dilution), you will need to determine the molar concentration. The Bradford estimation is not good enough. It would be more accurate to perform an amino acid analysis - http://www.sigmaaldrich.com/analytical-chromatography/analytical-reagents/amino-acid-analysis.html - which involves complete hydrolysis of the protein, separation and quantification of amino acids relative to a quantitative standard that is commercially available (http://www.sigmaaldrich.com/analytical-chromatography/analytical-reagents/amino-acid-analysis.html). From there, you would calculate how many micromoles or nanomoles you have of the rarest amino acid, that is stable to the acid hydrolysis. Using your knowledge of the protein molecular weight, you can do the calculation to arrive at the number of micromoles or nanomoles of protein you have per ml of solution. At this point, you would be able to apply the Beer-Lambert law because you will have both the absorbance at 280 nm and the molar protein concentration. 

The expasy site that Jiri mentions is a good one. Practically, you need to be sure that your protein is >95% pure. If you have a contaminant that has lots of tryptophans (or if your protein has none) your estimation will be off using the 280nM read off. Bradford is better in such a case. If Bradford and UV are within10-20% error range you have nothing to worry. 

Hope this helps

The theoretical calculation of extinction coefficient based on amino acid composition assumes that the protein is completely denatured in 6M guanidine hydrochloride.

To determine the extinction coefficient experimentally, just make a solution of the protein with a known concentration (it could be based on the Bradford assay result or some other measurement of concentration) and measure the absorbance with a spectrophotometer, noting down the pathlength of the cuvette, and selecting the wavelength with the highest absorbance (usually about 280 nm, but it could be a different wavelength if there is a bound cofactor, prosthetic group, or ligand). The extinction coefficient is the absorbance divided by the concentration and the pathlength, according to Beer's Law (epsilon = absorbance/concentration/pathlength). The units of extinction coefficients are usually M-1cm-1, but for proteins it is often more convenient to use (mg/ml)-1cm-1. If the molecular mass of the protein is known, divide mg/ml (-g/L) by mass in g/mole to get mole/L.

A few things to keep in mind:

The Bradford assay does not always give the true value of protein concentration because it depends to some extent on the amino acid composition, and your protein may have a different composition from the standard used in the assay (usually BSA). However, it can still be used for an operational measurement of concentration and extinction coefficient.

UV absorbance can be exaggerated by light scattering caused by aggregated protein, and also by anything bound to the protein that absorbs. Making the absorbance measurement in 6M guanidine HCl should take care of the aggregation problem, but won't help with bound ligands or prosthetic groups.

Thank you very much all Now i got idea

I think this link is very useful, 

http://www.instanano.com/characterization/theoretical/concentration-calculation-from-uv-vis-absorbance/

Try http://www.protparam.net. Will let you enter your protein sequence and OD280 reading and give you your protein concentration. Much faster than expasy prot param too

Molar Extinction Coefficients vs. Absorbances for 1% Solutions

Application of a molar extinction coefficient in the calculation yields an expression of concentration in terms of molarity:

A / εmolar = molar concentration

However, many sources, including the reference cited above, do not provide molar extinction coefficients. Instead, they

provide absorbance (A280nm) values for 1% (= 1g/100mL) solutions measured in a 1 cm cuvette. These values can be

understood as percent solution extinction coefficients (εpercent) having units of (g/100mL)

-1 cm-1 instead of M-1

cm-1

.

Consequently, when these values are applied as extinction coefficients in the general formula, the units for concentration, c,

are percent solution (i.e., 1% = 1g/100mL = 10mg/mL).

A / εpercent = percent concentration

If one wishes to report concentration in terms of mg/ml, then an adjustment factor of 10 must be made when using these

percent solution extinction coefficients (i.e., one must convert from 10 mg/ml units to 1 mg/ml concentration units).

(A / εpercent) 10 = concentration in mg/ml

The relationship between molar extinction coefficient (εmolar) and percent extinction coefficient (εpercent) is as follows:

(εmolar) 10 = (εpercent) × (molecular weight of protein)

Still other sources provide protein absorbance values for 0.1% (= mg/mL) solutions, as this unit of measure is more

convenient and common for protein work than percent solution. This variation in reporting style underscores the importance

of carefully reading stated values to be sure that the unit of measure is understood and applied correctly.

Examples

A. Proteins and Protein Mixtures with Unknown Extinction Coefficients

If no extinction coefficient information exists for a protein or protein mixture of interest, and a rough estimate of protein

concentration is required for a solution that has no other interfering substances, assume εpercent = 10. Most protein extinction

coefficients (εpercent) range from 4.0 to 24.0.5 Therefore, although any given protein can vary significantly from εpercent = 10,

the average for a mixture of many different proteins likely will be approximately 10.

B. Immunoglobulins

Most mammalian antibodies (i.e., immunoglobulins) have protein extinction coefficients (εpercent) in the range of 12 to 15.

Therefore, for typical antibody solutions, assume A 14 1%

280nm = or A A 1.4 1mg/ml

280 nm

0.1%

280nm = = .

For a typical IgG with MW = 150,000, this value corresponds to a molar extinction coefficient (ε) equal to 210,000M-1 cm-1

.

Pierce Biotechnology PO Box 117 (815) 968-0747 www.thermo.com/pierce

3747 N. Meridian Road Rockford, lL 61105 USA (815) 968-7316 fax

3

C. Bovine Serum Albumin (BSA)

Thermo ScientificTM PierceTM Albumin Standard Ampules (Product No. 23209) are provided as 2mg/mL solutions of purified

bovine serum albumin (BSA) in 0.9% NaCl. The product is calibrated by direct comparison of the absorbance at 280nm to a

known concentration of a BSA standard from the National Institute of Standards and Technology (NIST). Numerous values

for the absorptivity of BSA have been reported in the literature but are generally ~6.6 for a 1% solution at 280nm.

Therefore, the predicted absorbance at 280nm for the Albumin Standard, assuming exactly 2mg/mL and εpercent = 6.6 is

εpercent c L / 10 = A

[(6.6)(2.000)(1)] / 10 = 1.320

Suppose that relative to a water reference a researcher obtains a 280nm absorbance reading of 1.346 for the Albumin

Standard. The calculated concentration, assuming the stated percent absorptivity value, is as follows:

(A / εpercent) × 10 = cmg/ml

(1.346 / 6.6) × 10 = 2.039mg/mL

Assuming a MW = 66,400, the molar extinction coefficient at 280nm for BSA is approximately 43,824M-1 cm-1

.

Using Thermo Scientific Pierce Albumin Standards

If you plan to use a Pierce Albumin Solution (Product No. 23209 or 23210) as an absorbance standard, assume its

concentration to be accurate (that’s the whole point of a standard!) and use it to calculate your own “system-specific”

extinction coefficient. Do this by measuring the absorbance of the provided solution (or several dilutions thereof, ideally

prepared in duplicate or triplicate) and then applying the formula A / c L = ε. As in the above examples, this ε that you

calculate will be in terms of the units you used for c. The resulting “system-specific” extinction coefficient will be accurate

for your particular buffer, spectrophotometer and cuvette, etc., allowing the albumin to function as an accurate reference

standard for protein samples of unknown concentration. Without also knowing the extinction coefficient for the proteins in

the sample, you will not know whether the same concentrations of BSA and sample protein will have the same absorbance.

(For example, 1mg/mL IgG has nearly twice the absorbance of 1mg/mL BSA.) However, you will be able to use the BSA

standard as a uniform reference to compare and normalize multiple samples to each other.