Beyond thinking about the technical replicates in the ELISA assay, when you have your ELISA absorbance values of the standards and samples, you should draw a standard curve  to solve the concentrations of your unknown samples. I'll try to explain how to do this in Excel.

First you have to average the absorbance values from any duplicate determinations, then subtract the absorbance value of your blank sample (buffer only) from both the standard & sample absorbances. Then you should plot the blank-subtracted absorbances of the standards against their protein concentration (eg pg/ml). I do this with standard concentration on the y-axis and absorbance on the x-axis. By right-clicking the graph data points, choose "add a trendline" and choose one that best fits your concentration vs absorbance data. Typically in ELISA data, the best fit is a 2nd or 3rd order polynomial function. When choosing the trendline, tick the boxes "Display Equation on Chart" and "Display R-squared Value on Chart" from the same command box. The R squared value will aid you (together with the visual trendline) in choosing the best fit for your data; the closer to 1 the better. The equation displayed next to your graph in the forma y= .... will then give you the calculation for y= pg/ml based on x=absorbance value. Just feed the equation into a desired cell in excel, so that you replace all the "X"-values in the equation by the absorbance in your unknown sample.

For example, if you got the following 2nd order polynomial function for your standard curve:

y = 880,33 x^2 + 1979,3 x + 70,978 and the absorbance of your unknown sample is written in the cell A12 on your excel spread sheet, you would calculate the protein concentration in this sample by writing 

=880,33*A12*A12 + 1979,3*A12 + 70,978

and just copy the formula to all other unknown samples...

I hope this helps; don't know if this is the best way to plot your ELISA data, but it should be one valid option. 

Hi Ashraf,

Always run ELISA samples in duplicate or triplicate. This will provide enough data for statistical validation of the results. Many computer programs are now available to help process ELISA results. you can calculate on excel also...

hope this help

The same as for RIA. Always better to have tripicates for statistical analysis

Beyond thinking about the technical replicates in the ELISA assay, when you have your ELISA absorbance values of the standards and samples, you should draw a standard curve  to solve the concentrations of your unknown samples. I'll try to explain how to do this in Excel.

First you have to average the absorbance values from any duplicate determinations, then subtract the absorbance value of your blank sample (buffer only) from both the standard & sample absorbances. Then you should plot the blank-subtracted absorbances of the standards against their protein concentration (eg pg/ml). I do this with standard concentration on the y-axis and absorbance on the x-axis. By right-clicking the graph data points, choose "add a trendline" and choose one that best fits your concentration vs absorbance data. Typically in ELISA data, the best fit is a 2nd or 3rd order polynomial function. When choosing the trendline, tick the boxes "Display Equation on Chart" and "Display R-squared Value on Chart" from the same command box. The R squared value will aid you (together with the visual trendline) in choosing the best fit for your data; the closer to 1 the better. The equation displayed next to your graph in the forma y= .... will then give you the calculation for y= pg/ml based on x=absorbance value. Just feed the equation into a desired cell in excel, so that you replace all the "X"-values in the equation by the absorbance in your unknown sample.

For example, if you got the following 2nd order polynomial function for your standard curve:

y = 880,33 x^2 + 1979,3 x + 70,978 and the absorbance of your unknown sample is written in the cell A12 on your excel spread sheet, you would calculate the protein concentration in this sample by writing 

=880,33*A12*A12 + 1979,3*A12 + 70,978

and just copy the formula to all other unknown samples...

I hope this helps; don't know if this is the best way to plot your ELISA data, but it should be one valid option. 

Thank you all for valuable suggestion. Special thanks to Kristiina for nice detailed explanation.