there are various methids... the basics behind all the method is the 260/280 nm ration measurement for purity and 260/230 nm for salt contamination measurement. As Mashusudan singh rathore shared, nanodrop is one of the best instrument available from thermo scientific for measuring the DNA conc and purity using as low as 0.5ul of the sample.
Normal spectrophotometric measurements willalso be good if ou cary out precisely...
In order to measure DNA concentration and to guess whether the quality of your tested DNA is good or not (260/280) you have to use the Nanodrop. It gives you all proporties of the tested samples.
I agree in recommend Nanodrop. However, one limitation is, depending on the kind of investigation and on the sample type, it doesn't discriminates between different DNAs, e.g., human from non-human DNA in the same sample.
I think that the exact work and the ability to manipulate the size of extract makes it easy to deal with any device, for example photometry, this when the device that you are looking for is not available. Nanodrop can be the solution .
We have successful used invitrogen Qubit® 2.0 Fluorometer which gives the same results given by the 260nm UV readings obtained from a "clean" DNA (ethanol precipitated). If the DNA comes from a miniprep or other spin-column kits (where it is normally associated with residues from the buffers used for purification which absorb significantly in the UV) you cannot use UV-based systems because you always get values much higher than the real ones. Instead Qubit is ok also in this case.
You can verify this easily this way:
1) measure the DNA concentration of a sample purified by ethanol precipitation
2) purify this sample with spin columns (i.e. PCR purification kit)
3) elute with a volume equal to the starting sample volume
Now if you measure the concentration by UV it will be much higher than the original which is of course impossible!
With Qubit or possibly other systems similar this doesn't happen and this systems looks very reliable.
N.B. Only BR reagent (HS for very low concentration was not satisfactory instead) from Qubit system provided reproducible results in this way in our tests.
NanoDrop is not accurate for low DNA concentrations (below 20 ng/microL). For those we use QuantiFluor from Promega. Works pretty well, is almost as fast, but less expensive, as the NanoDrop device.
Spectrophotometric methods often over estimate the DNA conc. Flourescent molecules that bind to dsDNA is a well used method that is supposed to be more accurate. For instance "Qubit", "picogreen".