Do online and offline data collection methods yield different attribute preferences in an importance-rating task and a choice-based conjoint analysis task? This question is addressed in a field study designed to identify promising attributes for a new generation of wireless telephone handsets in an actual product development context. No practical differences in attribute preferences were observed between data collection methods when attribute preferences were measured using a direct importance-rating question. However, significant differences in attribute preferences were observed between the methods in the choice-based conjoint analysis task. Moreover, the online data collection method was judged superior to a traditional offline (paper-and-pencil) method on the basis of internal consistency and predictive (face) validity. These findings support the use of Internet/Web-enabled technology for conjoint analysis data collection. Other implications for research practice are discussed.