Full-day classroom observations of learning disabled, emotionally disabled, educable mentally retarded, and nonhandicapped students (N = 122) were conducted. A 10-second interval recording system was used to document tasks and student responding during written language instruction in regular and special education settings. Regardless of categorical designation, extreme variability was found in the amount of time individual students spent in writing activities or in receiving written language instruction. Implications for improving written language instruction for students with mild handicaps include increasing allocated time, teaching written language as an integrated process, and coordinating written language activities with different content areas. Collaborative problem solving between regular and special educators is needed to accomplish these changes.