The nine-hole box is commonly used to evaluate attention performance using a visual discrimination task in laboratory animals.

The nine-hole box is composed of a test chamber, food or drink dispenser, a Link Box to connect it to the PC and the PackWin software.

The nine-hole box is assembled with black aluminium walls and a transparent front door. The box is equipped with an arc of 9 contiguous apertures set into the rear wall, a house light, a food pellet dispenser and a `pusher` to detect the nose-pokes into the food holder. The holes not used in the experiment may be blocked up using a metal insert. Each hole is equipped with photocell beams and internal LED providing visual cues specific to each hole. The intensity of the LED can be adjusted in Link Box using the digital selector. The box is placed on a stainless-steel platform and the associated tray is easily removable to clean.

Panlab/Harvard Apparatus also offers an optimized nine-holes box for performing test in mice. This new box is supplied with 9 pellet dispensers in order to give the reward directly into the right stimulus hole when a correct response is fulfilled.

All Panlab/Harvard Apparatus nine-hole boxes are associated with the potent and versatile PackWin software in order to control the experiment (protocol configuration, experiment running) and obtain relevant data such as correct responses, incorrect responses, omissions, premature responses, perseverant responses, time out responses, total receptacle head entries, etc.

Different experimental paradigms for sustained attention, animal models of impulsive behavior and lateralized-discrimination task can be conducted using the nine-hole box.

As an example, in the 5-choice serial reaction time task, short-lasting stimuli are given in pseudo-randomized order in one of the holes of the cage (commonly, hole 1, 3, 5, 7 or 9). If the animal nose-pokes into the correct hole, a reinforcement (pellet) is given. If the animal nose-pokes into an incorrect hole, a time-out period (no light) is given and next trial begins. The choice accuracy (% of correct responses) gives an idea of the functional integrity of the attention as well as learning processes. These parameters are mostly altered in animal models of schizophrenia and Alzheimer diseases.