The value of the skills in an economy depends in part on the quality and quantity of the supply and in part on the demand for skills. In Ghana the pattern of job creation over the last decade has been one where non-farm self-employment and jobs in small firms have exploded in importance relative to jobs in the formal sector. While there has been extensive research on the returns to education, there has been much less on the returns to skills training. In Ghana there is a highly developed apprenticeship system where young men and women undertake sector specific training which is paid for usually by those responsible for the apprentice and which yields skills used primarily in the informal sector. In this paper we use a recent urban based household survey with detailed questions on the background, training and earnings of workers in both wage and selfemployment to ask how apprenticeship compares with other forms of training in terms of pay and employment outcomes. We show that apprenticeship is by far the most important institution providing training and is undertaken primarily by those with junior secondary school or lower levels of education. In comparing those with some form of training to those with none, training to be a nurse or teacher gives by far the highest return, a three fold increase in earnings. In contrast those who have done an apprenticeship earn significantly less than those with no training. Once an allowance is made for the level at which the apprentice enters the system there is evidence that for those with the lowest level of education apprenticeship does lead to a substantial increase in earnings, some 41 per cent. For those with higher levels of education it does not lead to any increase at all. In fact the point estimates imply a decrease. We argue that the reasons for undertaking an apprenticeship can be found, in part, in the powerful effect undertaking an apprenticeship has in increasing the probability of informal employment relative to having no job. In contrast to these outcomes for apprentices we show that training to be a nurse or teacher not only pays off in earnings but substantially shifts the probability of being in formal relative to informal employment. In summary these forms of training, one undertaken in the private sector and the other in the public sector, are associated with radically different outcomes for those receiving the training.