This thesis considers the existence of, and potential causes of, continuation and reversal effects in stock market returns. In the first part of the research, a timeseries approach is used to consider the profitability of momentum trading strategies on fourteen major stock market indices. Momentum trading strategies exploit continuation in returns, and evidence of significant profits to such strategies therefore implies the presence of continuation effects in the data samples. Significant losses to momentum strategies, on the other hand, are indicative of reversal effects in returns. This part of the research identifies continuation effects in stock index returns over periods of 1 trading day and 10 through 252 trading days. The second part of the research explores the various behavioural and nonbehavioural theories proposed in the literature for the existence of continuation and reversal effects in returns. Such effects imply that stock market trends differ systematically from trends in random data with the same underlying distribution of daily returns. An algorithm from the information technology literature is adapted and used to identify turning points in trends in the fourteen data sets, and the statistical properties of daily returns within stock market trends are analysed. Important patterns are observed in the steepness of trends and the volatility of returns within trends as they develop. These patterns enable some inferences to be drawn as to the most probable factors driving the continuation effects observed in the first part of the research, with nonsynchronous trading and investor loss aversion highlighted as potential causes of very short-term and medium-term effects respectively