Children with autism exhibit deficits in their quantity and quality of joint attention. Early autism intervention studies rarely document improvement in joint attention quality. The purpose of this study was to determine whether there was a change in joint attention quality for preschoolers with autism who were randomized to a joint attention intervention, symbolic play intervention, or a control group. Quality was defined as shared positive affect during joint attention as well as shared positive affect and utterances during joint attention. Interactions of group and time were found for both types of joint attention quality. During the follow up visits, the joint attention and symbolic play intervention groups produced more of these two types of joint attention quality than the control group.