The gender division of childcare has received less research attention than the division of paid work and unpaid household work, possibly because time for childcare has been presumed to have the same determinants as time for household work and because of lack of appropriate data. Further, some of the previous studies of parents' time for childcare have been subject to limitations, such as small sample size, selective samples, and self‐reports on childcare time, and few had access to earnings data. This paper uses register data on days of parental‐leave used by mothers and fathers of Swedish children born in 1994, including information on the earnings of mothers and fathers, to analyse the determinants of fathers' participation in childcare. In 1994 parents were entitled to 15 months of parental leave of which 12 months were compensated at 90 per cent of prior earnings. Our major finding is that while both father's and mother's earnings had positive effects on father's leave use, this was smaller at higher levels of earnings, and father's earnings had a greater impact than mother's. At similar levels of earnings, the more educated fathers used more leave. Fathers also used more leave if the mother had more schooling and if they were established in the labour market, but used less leave if the mother was established in the labour market.