Since many macroeconomic models are linear, it is not desirable to use themwith an asymmetric dependent variable. In this article, we formulate aunivariate test for symmetry, based on the third central moment and extendit to a multivariate test; the test does not require modelling and it is robustagainst serial correlation, Autoregressive Conditional Heteroscedasticity(ARCH) and nonnormality. In the empirical application of the test it isfound that orthodox theory seem to be supported; consumption expendi-ture on durable goods is found to be symmetric while consumptionexpenditure on nondurable goods is asymmetric for the USA and the UK,with peaks being higher than troughs are deep. Also, the empiricalimportance of the choice between the univariate and the multivariate testfor possibly correlated series is underscored; the results from the twoapproaches  clearly  differ.  Given  the  widespread  practice  of  usingconsumption expenditure on nondurable goods as the dependent variablein linear models for the USA and the UK, our results might be noteworthy.