AbstractIn the current investigation, we evaluated a method for increasing self‐feeding with 3 children with a history of food refusal. The children never (2 children) or rarely (1 child) self‐fed bites of food when the choice was between self‐feeding and escape from eating. When the choice was between self‐feeding 1 bite of food or being fed an identical bite of food, self‐feeding was low (2 children) or variable (1 child). Levels of self‐feeding increased for 2 children when the choice was between self‐feeding 1 bite of food or being fed multiple bites of the same food. For the 3rd child, self‐feeding increased when the choice was between self‐feeding 1 bite of food or being fed multiple bites of a less preferred food. The results showed that altering the contingencies associated with being fed increased the probability of self‐feeding, but the specific manipulations that produced self‐feeding were unique to each child.