Each year, the Australian Institute of Professional Photography conducts its National Print competition at which photographers from all over the country have the excellence of their images acknowledged by their peers through the allocation of Gold and Silver awards for the work determined to be the highest quality of that submitted for review. A panel of four or five respected commercial photographers make the judgments based on their experience and a set of prescribed standards. Detailed records of images, with corresponding judges' scores, have been collected since 2007, involving hundreds of images. Using these data, this research study was designed to determine what characteristics or qualities (features) of a photograph were the best predictors a high scores from judges. Furthermore, questions were asked regarding the characteristics of images likely to lead to variable responses from judges and those attributes of judges (e.g., sex, age, experience) that were associated with variability in scoring. Image features were identified through formal analysis based on principles and elements of design (concrete criteria) and through ratings of affective and communicative value (abstract criteria). Regression analyses were used to determine the importance of this range of variables in predicting the judges' scores. Comparisons also were made between the responses of professional photographers as judges and corresponding scores allocated by (a) a group of photographic educators and (b) a web-based computer image analysis system (the Aesthetic Quality Inference Engine). Discussion considered factors identified as contributing to photographic quality, and explored differences in effectiveness between criterion-based (algorithm dependent) and holistic aesthetic judgement.