A LITERATURE SEARCH YIELDED 35 EMPIRICAL STUDIES IN ORGANIZATION DEVELOPMENT FOR THE PERIOD 1959 TO MID-1975. EACH STUDY WAS ANALYZED IN TERMS OF RESEARCH DESIGN, DATA COLLECTION PROCEDURES, CHARACTERISTICS OF SUBJECTS, TREATMENT DIMENSIONS, AND APPROACHES TO DATA ANALYSIS. IT IS CONCLUDED THAT THE OVERALL QUALITY OF RESEARCH METHODOLOGY IN ORGANIZATION DEVELOPMENT HAS BEEN INCONSISTENT. RESEARCH DESIGNS HAVE BEEN FAIRLY STRONG, WITH MANY INVESTIGATORS RELYING ON QUASIEXPERIMENTAL DESIGNS. THERE HAS BEEN AN EXCESSIVE RELIANCE ON QUESTIONNAIRES AS THE SOLE APPROACH TO DATA COLLECTION. MOST STUDIES HAVE BEEN CONDUCTED IN ONE ORGANIZATION OR IN SITUATIONS IN WHICH THE NUMBERS OF UNITS OF ANALYSIS LARGER THAN THE INDIVIDUAL WERE SMALL (LESS THAN 10). MOST PROJECTS REPORT USING MORE THAN ONE INTERVENTION. LABORATORY TRAINING (BOTH PROCESS-ORIENTED AND TASK-ORIENTED) IS THE MOST COMMON INTERVENTION. THERE HAS BEEN A STRONG SHIFT FROM PROCESS TO TASK ORIENTATION IN MORE RECENT YEARS. OVER 75 PERCENT OF THE STUDIES REPORTED USING STATISTICAL TESTS OF SIGNIFICANCE. DATA ANALYSIS IS BECOMING MORE SOPHISTICATED, BUT THE VAST MAJORITY OF STUDIES HAVE USED VERY SIMPLE ANALYTICAL TECHNIQUES. RESEARCHERS IN ORGANIZATION DEVELOPMENT SHOULD DO THE FOLLOWING: MAKE GREATER USE OF QUASIEXPERIMENTAL DESIGNS AND THE MULTIPLE MEASUREMENT APPROACHES; FOCUS ON ORGANIZATION DEVELOPMENT PROGRAMS IN ORGANIZATIONS THAT CONTAIN LARGE NUMBERS OF HIGHLY COMPARABLE SUBSYSTEMS; STUDY ORGANIZATIONS OVER LONGER PERIODS OF TIME; INCREASE THE FREQUENCY WITH WHICH RESEARCH VARIABLES ARE MEASURED; MAKE DATA COLLECTION METHODS MORE ECLECTIC (BY INTERVIEWS, OBSERVATIONS, AND OTHER APPROACHES TO SUPPLEMENT QUESTIONNAIRES); AND MAKE ANALYTICAL PROCEDURES MORE MULTIVARIATE. A LIST OF REFERENCES IS INCLUDED. (AUTHOR ABSTRACT MODIFIED)