While saccade targeting in text reading is thought to be word based, in scene viewing research there is an ongoing debate between saliency-based targeting and object-based targeting. Robust Preferred Viewing Location (PVL) effect for words and significant influence of factors such as word length and launch site distance constitute a strong argument in favour of the word-based saccade targeting hypothesis in reading. Also, related effects – the refixation Optimal Viewing Position (OVP) and the fixation duration Inverted Optimal Viewing Position (IOVP) for words have been confirmed in reading studies. Following recent results reporting PVL for objects in scene viewing, the current study further investigated the phenomenon using a corpus of eye tracking data collected from 72 participants looking at the 135 naturalistic scenes with 730 tagged objects in aesthetic judgement and memorisation tasks. The PVL for objects was found to be influenced by both object size, and object centre based launch site distance, in the same direction as hypothesised based on the reading literature, i.e. large size of the object and long launch site distance shifted the distribution towards the object edge closest to the launch site. Although there was no evidence for refixation OVP effect, fixation duration IOVP was found for both first and single fixations. These findings contribute to the existing evidence in favour of object-based saccade targeting and point to the parallel between the way we move our eyes in reading and in scene viewing.