A three-year research effort was conducted to design a crime laboratory proficiency testing program encompassing the United States. The objectives were to: (1) determine the feasibility of preparation and distribution of different classes of physical evidence; (2) assess the accuracy of criminalistics laboratories in the processing of selected samples of physical evidence; (3) conduct statistical studies of the tests administered; and (4) establish the basis for the design of education and self-improvement programs which will assist the criminalistics profession to higher levels of proficiency. The following types of samples were manufactured, packaged and distributed: controlled substances, blood, paint, glass, hair, fibers, firearms, physiological fluids (semen, saliva), questioned documents, wood, arson accelerants, soils, and metals. The data were analyzed by giving full, partial or no credit, depending upon how thoroughly the work was performed. It was concluded that voluntary, anonymous proficiency testing is both feasible and necessary, and that there is a need for ongoing proficiency testing. A wide range of proficiency levels was found among the nation's laboratories, with several evidence types posing serious difficulties for the laboratories. The majority of laboratories lacked the financial resources to participate in a proficiency testing program on a fee basis. Several recommendations were made. (Author/GDC)