Small proteins found in plasma such as cytokines, chemokines and growth factors are involved in aging, disease pathogenesis, embryonic development, non-specific response to infection, specific response to antigens, and changes in cognitive functions. In addition, such analytes have roles in stem cell differentiation, vaccine efficacy and allograft rejection. Often the expression of these analytes are obtained from bead based multiplex immunoassays such as Luminex’s xMAP® technology.  Software systems such as xPONENT® from Luminex and Bio-Plex®  Manager™  output tabulated sheets containing observed concentration values, fluorescent values and standards and expected concentration values as well as background corrected values for test, controls, standards and blanks samples.  However, from such systems much of the life scientist’s test data can end up classified as being out-of-range leaving him or her with the problem of working with sparse and unbalanced data that in turn reduces their power of discovery and analysis. We have developed new approaches to the analysis of bead-based immunoassays that can reduce the number of out-of-range values by 50%.  We show that basing the analysis on raw and mean fluorescent values is often a better choice than relying on the derived concentrations values obtained from commercial software systems. We show how to analyse and to normalize the test data to compensate for plate-to-plate variations, how to check the reliability of test samples even if all or most of the concentration data is deemed out-of-range, and we show how to get improved estimates of analyte concentrations. In this study we use the expression of 14 analytes, using 6 commercially available kits, across 177 patients, recorded at two time points and the associated analyte standards, controls and blanks.  In total 60 micro titre plates are used resulting in 4965 readings. We show that the common practice of subtracting analyte blanks from test samples has no rational basis. We argue for our analysis of protein expression that the fluorescent values are a better choice than absolute concentration values. The significance of this work for the life scientist means higher statistical power and even lowers experimental costs.