Purpose: In an elderly Australian population, to evaluate the quality of fundus photographs taken non-mydriatically in both eyes, and to compare the quality of those taken second with those taken first. Methods: From 2258 participants (4516 images) aged 70 years and older who participated in the Melbourne Collaborative Cohort Study (MCCS), digital non-stereoscopic 45? retinal photographs were taken with a Canon CR6-45NM Non-mydriatic Retinal Camera and evaluated. The quality of macular images was assessed as good, fair, and poor and McNemar s test was used to analyze variation in quality. Results: Gradable quality images were obtained from 95.8 eyes of participants, with 93.9 of participants having gradable photos of both eyes. The gradable rate for the eye photographed first (right), was significantly higher than that for the eye photographed second (left): 89.7 vs. 85.6, respectively (difference of 4.12, confidence interval [CI] of 2.685.54, p <0.001). The rate of ungradable photographs from the second eye was slightly greater than the first eye (4.5 and 3.8, respectively), but the difference in proportion was not statistically significant (difference of 3.6, CI of 0.171.5, p = 0.384). Conclusions: In the setting of a large elderly cohort study, non-dilated 45? digital retinal imaging is an excellent method for fundus examination. It is fast, easy to use, noninvasive, and a reliable AMD (age-related macular degeneration)-detecting technique with only a minor loss of information from the second eye.