The "two-photon" effect is exploited in laser scanning confocal microscopy. Here is a good description.
1P vs 2P fluorescence imaging
The scanning blue (488nm) laser excites an entire column of sample.
The scanning IR pulse laser (Zeiss NLO system) excites only a small spot of sample.
The image was taken on a Zeiss LSM NLO system using a Nikon D1X digital camera. The sample is a dilute fluorescein solution in a quartz cuvette. No image manipulation was performed (except for creating the magnified insert).

Image by Steve Ruzin and Holly Aaron, UC Berkeley

The lateral resolution in WF fluorescence (based on Rayleigh's criterion) is 0.6(lambda-em)/NAobj. This relationship accounts for the blurring effect of the nth order diffraction rings around the Airy Disk. With Confocal the pinhole reduces or eliminates the rings resulting in a higher resolution: d=0.37(lambda-em)/NAobj. Note that resolution is Em-dependent.

With 2P however, it's a little confusing. While it is still true that resolution is dependent on the wavelength of emitted light, the size of the excitation spot is so much larger than its equivalent in 1P confocal (i.e., in 1P it is not the resolution-limiting factor, in 2P it is) that the constant in the resolution equation is greatly increased. What limits resolution in all microscopy is spot diameter, and that diameter is dependent on the longest wavelength. In 1P the longer wavelength is the Em light. In 2P the longest wavelength is the Ex light.

Thus for 2P resolution is described as d=0.7(lambda-em)/NAobj.