A modified version of the Rockport Fitness Walking Test (RFWT) has been determined to be reliable for persons with mental retardation. Previous investigations have used one pacer for each participant. While this protocol may be reliable, it is not likely feasible in most school or community-based settings. Therefore, the purpose of this study is twofold: first to evaluate the reliability of a modified version of the RFWT using one pacer per five participants; second, to determine the agreement among the one to one protocol and the proposed five to one protocol. This study compared the end completion times and mean peak heart rates of three different one-mile walking sessions of 35 participants with mild mental retardation. Eighteen females (21.1 �� 2.7 yr.) and 17 males (19.7 �� 2.8 yr.) participated in each testing session. Each participant randomly completed two sessions of walking with one pacer per five participants and one session of one pacer per participant on an indoor surface 1/9th of a mile long. Results from ANOVA indicated end completion times and mean peak heart rates did not differ based on gender (p=0.798), sessions (p=0.053), and gender x sessions (p=0.855). To determine agreement, intraclass correlation coefficients (ICC) were calculated for end completion times for the first five to one and the one to one (R=0.83) and the second five to one and the one to one (R=0.85). ICC's were calculated for mean peak heart rates for the first five to one and the one to one (R=0.95) and for the second five to one and the one to one (R=0.93). The results support that the five to one modified walking protocol produces results similar to the results using the one to one testing. / Graduation date: 1998