PROBLEMS ASSOCIATED WITH THE FORMULATION AND TESTING OF LABELING THEORY ARE HIGHLIGHTED. IT IS SUGGESTED THAT A MERGER OF LABELING AND CAUSAL ATTRIBUTION THEORY RESOLVES MANY OF THESE PROBLEMS AND FACILITATES OUR UNDERSTANDING OF DEVIANCE LABELING. A THEORY OF ATTRIBUTIONAL DIFFERENTIATION IS CONSTRUCTED TO EXPLAIN HOW RESOURCE VARIABLES OF OFFENDERS AND VICTIMS REGULATE THE DECISION TO LABEL ACTS AND ACTORS AS DEVIANT. PREDICTIONS ARE OFFERED AS TO THE IMPACT OF THE INTERACTANTS' RESOURCE VARIABLES ON THE ATTRIBUTION OF CAUSAL RESPONSIBILITY, AN ATTRIBUTION THAT IS A NECESSARY PREDECESSOR TO DEVIANCE LABELING. A PARTIAL TEST OF THE THEORY IS CARRIED OUT EMPLOYING 30 EXPERIMENTAL CONDITIONS, WITH SUBJECTS RANDOMLY ASSIGNED TO CONDITIONS. THE PRIMARY CRITERION VARIABLE IN THE STUDY SOUGHT TO MEASURE THE SUBJECTS' PERCEPTION OF HOW RESPONSIBLE EACH ACTOR (OFFENDER AND VICTIM) WAS FOR CAUSING THE CRIMINAL ACT DESCRIBED IN THE QUESTIONNAIRE. IN GENERAL, IT IS CONCLUDED THAT 'RESOURCE VARIABLES' (I.E., PERCEIVED STATUS AND BEHAVIORAL CHARACTERISTICS OFTEN IN THE FORM OF STEREOTYPES) MAKE A DIFFERENCE IN THE PERCEIVER'S ATTRIBUTION OF CAUSAL RESPONSIBILITY. MORE SPECIFICALLY, THE PREDICTIONS ARE SUPPORTED THAT ASCRIBED (E.G., SEX, AGE, RACE) AND ACHIEVED STATUS CHARACTERISTICS HAVE COMPETING EFFECTS AND THAT PREVIOUSLY EXISTING DEVIANCE LABELS HAVE THE MOST PROFOUND IMPACT ON ATTRIBUTIONS OF CAUSAL RESPONSIBILITY. SUGGESTIONS FOR EXPANSION AND UTILIZATION OF A GENERAL ATTRIBUTION-LABELING THEORY ARE OFFERED. BECAUSE OF THE MEASURES USED, THE PRIMARY STATISTICAL TOOL USED IN THE STUDY WAS THE KOLMOGOROV-SMIRNOV TEST. TABLES AND A LIST OF REFERENCES ARE PROVIDED. PRETEST DATA ARE APPENDED. (AUTHOR ABSTRACT MODIFIED--MLC)