Parallel merit reliability (PMR) indexes the same consistency of measurement that is reflected in a validity coefficient; it reflects the reliability of measurement across identical merit score cases. Research has identified the potential benefits of the PMR approach as providing item level and cut-score reliability indices without requiring multiple administrations of the same test. Disadvantages of this type of reliability include: the need for a set of merit scores in addition to the item or interval scores; low magnitude coefficients; coefficients that are sensitive to the scale of the merit score; and the potential for artifactual bias when the merit intervals are too broad or of unequal discrimination. Computation of a parallel merit coefficient can be problematic if merit interval means are used as true score estimates. Means may be non-linear, single-case intervals may inflate the variance of means, and data processing may be awkward. However, by assuming linear relations between merit and item score, use of a regression approach may be justified, thereby, avoiding some problems associated with the use of interval means. A five-step process was used to investigate this approach to PMR. Two types of reliability coefficients were computed--parallel merit and coefficients representing coefficients of stability--for a real data set (high school graduation competency test data for 319 students in the 1988-89 school year) and a simulated data set (1,000 abilities generated to respond to 250 items ranging from the least to greatest ability level). Under the procedures presented, merit interval means were often non-linear and non-representative due to the number of interval cases. This preliminary effort to explore the usefulness of PMR lacked a credible set of real data. Future analysis should include a real data set having cases that match on pretest, posttest, and achievement test scores. Four data tables and three graphs are provided. (Author/RLC)