Two studies assessed the reliability and utility of the Stimulus Preference Coding System (SPCS) to measure approach, avoidance, and happy and unhappy behaviors in persons with developmental disorders. Study 1 took place in an institutional setting. The nine participants were all adults with mental retardation and multiple associated disabilities. Inter-observer reliability ranged from 72% to 100%. Study 2 took place in an after-school setting. The four participants were children diagnosed with autism and mental retardation. Inter-observer agreement ranged from 70% to 91%. Approach and avoidance behaviors were a function of staff person and task. The SPCS may be useful in identifying reinforcers, promoting happiness, analyzing task and staff effects, and clarifying the relationship of stimulus preference to psychopathology.