Low-flow electrospray ionization is typically a purely electrostatic method, used without supporting sheath-gas nebulization. Complex spray morphology results from a large number of possible spray emission modes. Spray morphology may assume the optimal Taylor cone-jet spray mode under equilibrium conditions. When coupling to nanobore gradient elution chromatography, however, stability of the Taylor cone-jet spray mode is compromised by the gradient of mobile phase physiochemical properties. The common spray modes for aqueous/organic mobile phases were characterized using orthogonal (strobed illumination) transmitted light and (continuous illumination) scattered light imaging. Correlation of image sets from these complementary illumination methods provides the basis for spray mode identification using qualitative and quantitative image analysis. An automated feedback-controlled electrospray source was developed on a computer capable of controlling electrospray potential using an image-processing based algorithm for spray mode identification. The implementation of the feedback loop results in a system that is both self-starting and self-tuning for a specific spray mode or modes. Thus, changes in mobile phase composition and/or flow rate are compensated in real-time and the source is maintained in the cone-jet or pulsed cone-jet spray modes.