A video-computer monitoring system was developed for use evaluating the effects of environmental factors and toxicants on fish behavior. The system recorded the X and Y coordinates of individual fish at 4-second intervals in a 50 cm square observation area and subsequently calculated 13 different behavioral parameters from the resulting time series of coordinates. The standard (control) swimming behavior of bluegills, Lepomis macrochirus, Raf., in the system was characterized in terms of activity, turning and spatial selection parameters over periods of 3 to 4 days and at two temperatures, 5 and 22 °C. Activity started at high levels on Day 1 and gradually decreased with time. The rate of this decrease was greater at 5 °C than at 22 °C. Turning behavior did not change in scale with time but did become more variable after an initial exploratory period. Spatial selection trends included a tendency for bluegills to spend more time around the periphery of the observation area than in the central area, and a tendency to spend more time in the downstream end of the observation area than in the upstream end. The behavioral responses of bluegills to abrupt, sub lethal (1 and 10% of the 96-hr LC50) gradients of ammonium chloride at 22 °C were assessed using the same system. The lower concentration gradients produced temporary increased activity and increased use of near-gradient areas. These responses were considered to be related to exploratory behavior. The higher gradients produced avoidance by 3 out of 4 fish; but these results conflicted with those obtained from previous tests conducted at 14 °C. Because of this discrepancy it was hypothesized that low temperatures interfere with the ability of bluegills to either detect ammonia after a certain period of exposure or their ability to behaviorally compensate for physiological stress produced by ammonia- gradients of this strength. / Ph. D.