Is it possible to measure the Absorbance at 280nm in a nanodrop or Trinean dropsense with your samples? THis may help to get a rough idea of your protein concentration. Of course it depends on what kind of samples you have and if there are contaminants also absorbing at this WL. Regards.

If you have an absorbance plate reader with UV capability at 280 nm, you could put a fixed dilution of each extract in a 96-well clear acrylic plate (the kind that allows absorbance to be measured at such a short wavelength) or a quartz plate and measure the absorbance of all the samples at the same time. Once you determine the correlation between the 280 nm absorbance and the protein concentration from Bradford, you will have a good estimate of the dilution you need to make for each sample. You could also make 280/260 absorbance measurements this way.

In terms of time, dilution trial & error with Bradford reagent as well as determining the correlation with Abs at 280 would both involve  expenditure of time. However, the need for accuracy would get a higher priority  than the method taking a little extra time for execution.

Alternately, for a particular batch of Bradford reagent that you have prepared/obtained pre-prepared, you could prepare a standard plot of BSA. This would set the limits for linearity (set up adequate number of duplicates for an accurate average) of absorbance at 595 nm.

Thereafter dilute your protein sample 1:10, 1;100 & 1:1000 & add the same Bradford reagent & read the absorbance at 595 nm- whichever falls within the linear range , work with that concentration. You need not repeat the entire set of BSA concns. for the standard plot  every time you need to estimate protein, just keep 1-2 standard concns. in case you are doubtful.

As Dr.Schlib rightly pointed out 280 nm is OK provided that no contaminants would interfere at that wavelength (particularly in plant extracts, phenolics can be a problem for both estimation at 280 nm as well as for the Lowry's method). Also, cant the Bradford assay with your different protein concentrations be carried out simultaneously on the ELISA plate as indicated by Dr.Shapiro?

I agree with Mrinalini Menon that using a 96-well plate for the Bradford will be very helpful, if you aren't doing it already. An ordinary clear polystyrene plate, preferably with flat-bottom wells, can be used. An 8- or 12-channel pipettor will also be a great help, allowing you to add the reagent to many wells at the same time. I usually use a 10-µl sample volume and 200 µl of 1X Bradford reagent, mix on a plate shaker, wait 5 minutes, then read the absorbance.

For getting the good sample dilution?

First, go for 2 fold dilutions. And use the reagent: sample ratio either 1:4 or 1:3 and take the OD. Note down the sample dilutions which show OD around 0.5-0.6. Then from that particular dilution make further dilitions (1:5 or 1:10) and add reagent in the same above said ratios and take OD. Now you can get the accuarte rsults.

I would suggestion if you need to dilute sample then just don't use Bradford method since Bradford method is non specific. Second thing, we have observed that after dilution you won't get exact protein since the relation is not linear. But it is fast and easy and you can use 96 well plate