A maximum likelihood based iterative algorithm adapted from nuclear medicine imaging for noncoherent optical imaging was presented in a previous publication with some initial computer-simulation testing. This algorithm is identical in form to that previously derived in a different way by W. H. Richardson , “ Bayesian-Based Iterative Method of Image Restoration,” J. Opt. Soc. Am.62, 55– 59 ( 1972) and L. B. Lucy , “ An Iterative Technique for the Rectification of Observed Distributions,” Astron. J.79, 745– 765 ( 1974). Foreseen applications include superresolution and 3-D fluorescence microscopy. This paper presents further simulation testing of this algorithm and a preliminary experiment with a defocused camera. The simulations show quantified resolution improvement as a function of iteration number, and they show qualitatively the trend in limitations on restored resolution when noise is present in the data. Also shown are results of a simulation in restoring missing-cone information for 3-D imaging. Conclusions are in support of the feasibility of using these methods with real systems, while computational cost and timing estimates indicate that it should be realistic to implement these methods. It is suggested in the Appendix that future extensions to the maximum likelihood based derivation of this algorithm will address some of the limitations that are experienced with the nonextended form of the algorithm presented here.