Breeding testosterone (T) profiles of free-living male birds are hypothesized to reflect a trade-off between investment in competitive behaviour for mates or territories, typically accompanied by high T-levels, and investment in paternal care, typically accompanied by low T-levels. To test this hypothesis we monitored song activity, as a measure of territorial advertisement or mate attraction, and feeding efforts, as a measure of paternal care, in great tit Parus major males that either received T-filled or empty implants in the middle of the feeding phase, a period of high paternal commitment and low T-levels. In support of the trade-off hypothesis, T-implanted males sang significantly more than controls. However, we found no significant effect of the elevated T-levels on male and female feeding behaviour despite the large increase in T following implantation. Also, no short-term effects on male fitness were found. Taken together these results do not support the trade-off hypothesis. We discuss that the low overall responsiveness in our study might be related to a high fitness cost of neglecting offspring in favour of T-enhanced behaviour.