This paper examines the temporal relationship between population growth and economic growth for a sample of less developed countries. Unlike prior research efforts, Granger causality tests are conducted in the context of error correction models when cointegration is present. Of the thirteen countries examined cointegration was present in only three countries. Although ten countries did not exhibit properties of cointegration, it is noted that researchers undertaking time series studies of the relationship between population growth and economic growth utilizing differenced data should evaluate the possible long-run relationship. Capturing both the short-run and long-run behaviour of the respective time series may provide the researcher with a more robust test of Granger-casuality.