This study explores the effects of call attempts and time periods on call outcomes and sample composition. A mobile computer-assisted telephone interview survey was conducted to collect data from adult mobile phone users about use and attitudes towards mobile phones; paradata regarding call dispositions, time and day of the week of calls and number of call attempts was also available. The first call contact rate was approximately 27% and varied significantly across time periods; the rate fell to below 20% for the second call. Weekend time periods yielded higher contact rates than weekday time periods. The interview rate on the first call was 12% and decreased steadily in subsequent calls. Mobile phone numbers that yielded call rejection, voicemail or were busy on the first call were very difficult to convert into interview on the second call. The number of call attempts and time period of the calls affect sample composition, namely in relation to respondents' age, educational level and area of residence. Future research and practical implications of the findings for mobile CATI surveys are discussed.