Power? Wavelength? Circular vs Elliptical beam? I know that there are fluorophore considerations, but isn't there a range of different fluorophores available at different wavelengths?
1. If you do not absolutely need laser for your measurements, you might want to use a bright lamp to avoid ringing, speckles and other coherent headaches (I built both types of systems and you do need to have really good reason to go into coherent illumination). That, of course has nothing to do with cases where you are not using laser light directly to build your image, i.e. you use say fluorescence, and you do have rather good filters to suppress your laser light down to 10^(-6) at least.
2. I would suggest to build your entire illumination system as achromatic as possible.You might want to use several lasers for different chromophors, and you do want as easy switch.
I agree with Peter, UV requires rather special treatment, and is very limiting in optics choice.
Probably very expansive, but an ultimate solution would be femtosecond system with OPA.