In many biological applications, it is becoming increasingly important to extend imaging capabilities to encompass the simultaneous identification of multiple molecular species in order to gain a better understanding of the structural and functional elements of cells. A key element in this quest is the ability to rationally optimize various labeling techniques, which in turn requires an understanding of the basic physical processes operating during the labeling process, and the interactions among these basic processes. We have developed a mathematical model of the immunocolloid labeling process which in its most simplified form involves just two physical processes: diffusive transport of the labels and a surface-volume reaction modeling the binding of the labels in solution to the target sites. Even this simplified model contains parameters which are not easily accessible, but depending on which of the processes is the rate limiting step in the labeling process, we are able to use asymptotic analysis to obtain different approximate solutions which suggest how to use a relatively small number of experiments to determine which process is rate limiting and possibly extract estimates of the needed parameters. The model allows us to determine the time scale of approach to the steady state, which is of great interest in applications of the labeling process. We will also show how the simple asymptotic solutions mentioned above are relevant in the analysis of problems with more complicated geometry and a nonhomogeneous distribution of sites to be labeled.