Use of UV dyes is advantageous in many fluorescence detection applications due to the favorable dye properties (OPA, for example, tags the primary amines with a fast reaction at room temperature). However, UV dyes can only be excited with weak plasma sources or bulky and unreliable lasers. What is needed for this market is a rugged fluorescence detector with the sensitivity of the laser by the size of the plasma source. We propose to develop such a detector based on diode laser technology. Our calculations show it will be orders of magnitude more sensitive than conventional plasma-based detectors, yet significantly smaller and cheaper than the existing lasers. Our solid-state approach has no moving parts, and promises a mean time before failure of tens of thousands of hours. In the Phase I program, we will construct a laboratory model of the light source and measure the generated UV power. In Phase II, we will construct the detector, install it in a commercial separation or analysis system, and compare its performance with the best alternatives.