A study of the binocular accommodation response is presented for normal and amblyopic observers to selected stimulus conditions using a binocular infra-red optometer and a commercially-available autorefractor. The work reviews the neural control of the near triad and discusses the historical development of models of mutual interaction between accommodation and convergence, presenting experimental evidence to support or refute each proposition. The basic characteristics of the accommodation response are reviewed along with the influencing factors. A central feature of this work is the evaluation of the correlation present between the eyes for both step-wise changes in target vergence and steady-state viewing. Reaction times for visually normal subjects were found to be similar to those found by previous workers and were independent of both size and direction of the step change. Response times for a mean step size of approximately 2.5D exhibited a marked degree of intersubject variability, particularly for the decreasing response and were step-size dependent. Eye dominancy was not found to be a significant factor in the overall response time. The binocular accommodation responses were found to have a high level of correlation to step-wise changes in target vergence. This in, itself, is perhaps not surprising in view of the anatomical similarities between the eyes and the relatively large dioptric changes induced. To obtain a clearer picture of the control of accommodation assessment of the microfluctuations was necessary. A high degree of correlation between amblyopic eyes and their fellow normal eyes is reported for both reaction and response times, although response times are longer than those for normal eyes. Reaction times for four subjects were not significantly different to those of the dominant eye. The subject presenting with the deepest amblyopia did have a significantly increased reaction time and a relationship with minimum angle of resolution is considered. Steady-state viewing shows the microfluctuations to have a high level of coherence, suggesting the control of accommodation to be at or above the point at which the IIIrd nerves are conjoint. Increasing target vergence causes an increase in the rms amplitude of the microfluctuations, binocular viewing not influencing the response characteristics. As target luminance decreased, rms values and low frequency drifts increased. Amblyopic eyes show an increase in the magnitude of the low frequency components of the microfluctuations for moderate to high stimulus vergences. The presence of different behaviour to that observed in normals supports a role for the microfluctuations. The response of amblyopic eyes to coloured stimuli results in an increase of the low frequency component to targets at the extremities of the visual spectrum, furthering the argument for a positive role for the fluctuations. The steady-state response to coloured stimuli differed from that found in normal eyes in that the appropriate response to overcome the chromatic interval was not observed for moderate to high stimulus vergences. The overall anomalous response could not use the additional information provided by coloured targets. Finally the detectibility of defocus was tested with sine waves and using signals derived from the microfluctuations. The threshold of detection for the microfluctuations is similar to that for sine waves, but is thought to be due to the presence of discontinuities and abrupt shifts in the response level. This adds support to Crane's(1966) hypothesis of 'accommodative saccades'.