The serial correlation effects which non-synchronous trading can induce in financial data have been documented by various researchers. In this paper we investigate non-synchronous trading effects in terms of the predictability that may be induced in the values of stock indices. This analysis is applied to emerging-market data, on the grounds that such markets might be less liquid and thus prone to a higher degree of non- synchronous trading. We use both a daily data set and a higher frequency one, since the latter is a prerequisite for capturing intra-day variations in trading activity. When considering one-minute interval data, we obtain clear evidence of predictability between indices with different degrees of non-synchronous trading. We then propose a simple test to infer whether such predictability is mainly attributable to non- synchronous trading or an actual delayed adjustment on part of traders. The results obtained from an intra-day analysis suggest that the former cause seems a better explanation for the observed predictability. Future research in this area is needed to shed light on the degree of data predictability which may be exclusively attributed to non-synchronous trading, and how empirical results may be influenced by the chosen data frequency.