A new method for determining molecular weight distributions is explored. Matrix fluorescence photobleaching recovery (MFPR) requires dye attachment to the polymer analyte. No physical separation of macromolecules is accomplished, as in the predominant gel permeation chromatography technique. Rather, the distribution of diffusion coefficients is determined by inverse Laplace transformation of fluorescence photobleaching recovery decay profiles, and the diffusion data are then mapped onto molecular weight using a measured calibration. Resolution is enhanced by forcing the labeled polymer analyte to diffuse through a solution that contains unlabeled matrix, which can even be the same type of polymer. Relative to free diffusion, this increases the dependence of diffusion on molecular weight, making it easier to differentiate components having similar mass. Simulated, noise-corrupted data for molecules diffusing by reptation suggest that excellent resolution may be achieved. Empirical tests were conducted using fluorescent dextran and pullulan diffusers in unlabeled dextran matrix solutions. Although the resolution fell short of the simulated results, it exceeded that available from gel permeation chromatography as normally practiced. An investigation conducted to understand why the full resolution was not achieved revealed that matrix solutions of high molecular weight dextran lie in the unentangled semidilute regime, as evidenced by the scaling behavior of probe diffusion with probe molecular weight, the absence of a rheological plateau modulus, and the shape of the small-angle X-ray scattering profiles. Branching of high-molecular-weight dextran seems to preclude entry into the entangled regime, where a stronger diffusion vs mass relationship is expected. The good performance that was nevertheless achieved suggests that further exploration may prove fruitful.