The Lowry method for protein concentration is very repeatable. It is susceptible to interference by reducing agents such as 2-mercapthoethanol, or DTT. It is not susceptible to interference by detergents. Protein concentrations are determined using a standard curve as compared to a known sample, such as BSA.
The Bradford assay is essentially a dye binding assay. It is susceptible to detergents, but reducing agents have no consequence on this assay. I find this assay to have a poor linear response. This method also requires a standard curve vs a known sample.
UV assays are very good as long as there are no UV absorbing compounds in your preparation. Examples of UV absorbing compounds include nucleic acids and certain protease inhibitors such as benzamidine. In addition, any aggregates in your prep will make the assay difficult to read. Baring interference, the extinction coefficient of a purified protein can be calculated, making UV superior for determining the exact concentration. No standard curve is required.
