A preliminary study measured the contrast sensitivity function (CSF) in 30 children (aged 3 months-5 years). Preferential looking techniques were used to assess CSF to sine wave gratings displayed on one of two screens. To find a meaningful contrast sensitivity procedure we compared the results with a shorter procedure using an edge stimulus. The following problems were encountered: measuring the contrasts required to detect four or five different spatial frequencies took time, resulting in boredom and loss of attention in our subjects; there was poor correlation between CSF and edge detection; an interesting artefact resulted in a plateau rather than a low frequency fall-off in the CSF of five of the children greater than 30 months old. This artefact may have resulted from peripheral rather than central retinal responses and/or motion artefacts in the stimulus onset. A follow up study with 41 additional children aged 3-36 months limited the contrast testing to that of the spatial frequency corresponding to the peak of the CSF. The shortened procedure, plus a lot of encouragement, resulted in higher contrast sensitivities in all but the oldest age group and successful monocular contrast measurements. In order to avoid artefacts arising from peripheral vision, children were encouraged to look at each screen before responding.