Rogge and Fincham developed optimized measures of positive and negative relationship quality using a combination of exploratory factor analyses and IRT in a sample of over 1,600 college students. The authors asked respondents to rate their relationships on separate sets of 20 positive (e.g., enjoyable, pleasant, alive) and 20 negative (e.g., bad, empty, lifeless) adjectives, giving similar instructions to those used by Fincham and Linfield (1997; e.g., “considering only the positive qualities of your relationship and ignoring the negative ones, evaluate your relationship on the following qualities”). Factor analyses supported two dimensions of evaluation that were moderately correlated with one another. IRT analyses were used to identify the items most effective for assessing positive qualities (PRQ) and the items most effective for assessing negative qualities (NRQ). Hierarchical regression analyses showed that the PRQ-4 and NRQ-4 offered unique information beyond a 4-item measure of global relationship quality (CSI-4) for understanding self-reported positive interactions, negative interactions, satisfaction with sacrifice, vengefulness toward partner, hostile conflict behavior, and disagreement tolerance. Furthermore, the PRQ-4 and NRQ-4 displayed distinct patterns of validity within those regressions, with the NRQ-4 being more strongly related to things like vengefulness and hostile conflict behavior and the PRQ-4 being more strongly related to satisfaction with sacrifice and disagreement tolerance.