Quantitative behavioral assessment of captive cetaceans can provide significant insights into the physiological and emotional condition of individual animals.  The value of such assessments is highly dependant upon the proper design of the data collection protocol and its careful implementation.  It is also dependant upon the strength of the interpretation of the data.  The most powerful designs are those which characterize the external environment, physiological dynamics, and behavioral patterns concurrently.  This paper will discuss the design of behavioral observation protocols for captive marine mammals which allow correlation with environmental and physiological variables.  The specific problems of handling large quantities of data efficiently, and the factor of observer variability will be discussed.  The use of this type of protocol in conjunction with routine physical examinations to assess health changes will be emphasized.