To the Editor.— The article "Nonsteroidal Anti-inflammatory Agents in Rheumatoid Arthritis and Ankylosing Spondylitis" by Wasner et al (1981;236:2168) is an important, welladministered, and distinctive contribution to drug trials. I was educated. The experimental design is an important feature conducted in the "natural" setting of clinical practice, using six drugs as it usually happens in the physician's office. The data are useful for the average practicing physician or rheumatologist and are statistically more sound than those of most such studies. However, being an enthusiastic, long-term proponent of aspirin therapy in rheumatic disease, I offer the following criticism and suggestions: (1) There remains doubt that entericcoated aspirin are absorbed with enough predictability to make it worthwhile to use them. A better choice to assess aspirin is to take the drug as plain aspirin, crushed or powdered, at midmeal and at bedtime for a total of ±12 daily (1975; 232:1007, 1974;230:1385). 1-5