This paper presents an attempt to examine the applicability of the relative income hypothesis (RIH) in terms of its various specifications proposed by Duesenberry, Duesenberry, Eckstein and Fromm (DEF), Davis and the authors (MD). Using the time series data for 1951 through 1968 the analysis has been carried out for Canada, Finland, Guatemala, Honduras, India, Japan, Philippines, Sweden, United Kingdom and the United States. It is found that RIH provides a fairly good representation of the consumption behaviour of all the countries included in the study. All specifications, however, do not perform equally well. DEF and Davis functions score the maximum points; MD comes at par with DEF in case of Finland, Guatemala, and India. The original Duesenberry specification performs very poorly. This lends us to conclude that the process of habit formation is continuous contrary to what is implied by Duesenberry's original specification and that consumption is a better indicator of the standard of living than income is. Estimates of the long-run marginal propensities to consume are essentially the same as those computed from the permanent income hypothesis by Singh and Drost [1970]. This lends support to the view that the two hypotheses have essentially the same long-run implications. [ABSTRACT FROM AUTHOR]