Since the mid 1990-ies web survey data collection has been rapidly growing. The biggest advantage of this survey method is that it is easier, faster and cheaper than the ˇconventional”methods. However, quality issues in web surveys are an interesting theme to investigate. First of all, the complete coverage of the list depends on the available e-mail list that is an essential requirement for the first contact with the target population. Secondly, all statistical units need access to the Internet. Finally, improving the survey process quality in terms of timeliness of the information and in terms of response rate is a basic requirement for good product quality. In this study, data collection deals with university-to-work transition that has a great relevance in Italy given the high levels of young unemployment. Therefore, the target population is composed of graduate students. This population is very interesting for web survey testing since young generations usually have more familiarity with new technologies, such as the Internet. Moreover, in this particular context, the degree of list coverage is very high, given that the sample is composed of all respondents to a previews survey wave conducted by telephone interview in 2003. The aim of this paper is to analyse several dimensions of data collection mode quality by web adopting a mixed-mode contact strategy using both telephone and web. In such a way, it is possible to compare difference in response pattern between web and telephone forms on the occasion of both the first contact and the call-backs. Furthermore, we built different quality indicators that refer to timeliness of data collection (including the response speed at the first contact), and to the relationship between structural variables and the data-collection process quality.