The participants were all suspects of at least one criminal offense and were examined from January through March 1990. They were all United States citizens. Thirty-seven examiners conducted the polygraph examinations using field standard five-channel Lafayette and Stoelting polygraphs that recorded respiration, skin resistance, and the cardiovascular system. The numbers test took place between the first and second tests of the main test series in 251 cases; this procedure was the standard one. The numbers test took place before the first test of the main test series in the other 231 cases. Results suggested that the location of the numbers test had no effect on the rate of decisions regarding inconclusive results, deception indicated, and non-deception indicated. Findings indicated that the location of the numbers test could reasonably be left to the examiner's discretion, although several factors limit the generalizability of these results. Tables and 13 references