Despite the prevalence of microfluidic-based heterogeneous immunoassays (where analytes in solution are captured on a solid surface functionalized with a capture molecule), there is incomplete understanding of how assay parameters influence the amount of captured analytes. This study presents computational results and corresponding experimental binding assays in which the capture of analytes is studied under variations in both mass transfer and surface binding, constrained by real-world assay conditions of finite sample volume, assay time, and capture area. Our results identify: 1) a "reagent-limited" regime which exists only under the constraints of finite sample volume and assay time; 2) a critical flow rate (e.g. 0.5 microL min(-1) under our assay conditions) to gain the maximum signal with the fastest assay time; 3) an increase in signal by using a short concentrated plug (e.g. 5 microL, 100 nM) rather than a long dilute plug (e.g. 50 microL, 10 nM) of sample; 4) the possibility of spending a considerable fraction of the assay time out of the reaction-limited regime. Overall, an improved understanding of fundamental physical processes may be particularly beneficial for the design of point-of-care assays, where volumes of reagents and available samples are limited, and the desired time-to-result short.