Web surveys are the most successful form of Internet research and, particularly in the US, are becoming a mainstream market research tool. However, there is relatively little research on the effect of the web mode compared with more traditional administration methods. We review the literature on web surveys and sources of error, and investigate the impact of the web administration mode. Web and paper surveys were conducted simultaneously using near-identical self-completion questionnaires. A detailed comparison of responses included analysis of the means, skewness, kurtosis, and variance. After identifying a number of significant differences, chi-square and regression were used to analyse the administration mode effect. We argue that the differences found are attributable to sample bias rather than mode, and that the ramifications of low response rates in web surveys are more far reaching than has been recognised.