Previous studies have revealed that children increase the size of drawings of topics about which they feel positively and use their most preferred colours for colouring in these drawings, and decrease the size of drawings of topics about which they feel negatively and use their least preferred colours for colouring in these drawings. However, these previous findings have been obtained in studies employing drawing tasks where planning and production difficulties have been minimized by asking the children either to copy or to colour in an outline stimulus of a figure provided by the experimenter. The present experiment was designed to examine whether children also alter the use of size and colour in their drawings in response to emotional characterizations when they are not constrained by the presence of a model. In all 253 children aged between four and 11 years were asked to produce drawings of a neutrally, a positively and a negatively characterized topic (either a man, a dog or a tree). It was found that the children consistently increased the size of the positively characterized figures, did not consistently decrease the size of the negatively characterized figures, used their most preferred colours for the positive figures, and used their least preferred colours for the negative figures. These findings are discussed in relation to the operation of an appetitive-defensive mechanism and pictorial conventions.