Small-scale and cottage industries (SSMI) have received a considerable amount of attention in development strategies and policies in many countries. One of the main arguments in favour of small scale production is its potential to create employment, as it uses more labour per unit of output. Although the empirical evidence in the debate around the allocative efficiency of smallscale production is not unambiguous, scale-biased (macro) economic environments favouring large industries have been used as a justification for the promotion of small industries through direct assistence programmes.The effectiveness of small industry promotion programmes has received mixed judgements. One the one hand, (government) assistance has been praised as having ..a favourable effect on employment.. in small enterprises (Pernia and Pernia, 1986), while, according to others (UNDP et al., 1988), it may present in many cases an ..ineffective palliative.. for unfavourable demandconditions of (rural) SSMI. These evaluations are, however, not entirely comparable, since the latter one is an ..overall, qualitative assessment.. using effectiveness criteria such as outreach,assimilability, impact, sustainability, while the former one is based on a quantitative analysis of economic and social impact. This analysis of impact can thus be seen as part of a broad policy evaluation.This paper attempts to contribute to the debate around the impact of small industry promotion, using evidence concerning small industry assistence programmes in Central-Java, Indonesia. The analysis is based on the data set employed in Sandee et al. (1994); the selection of data and the methods used are different, however.In the next section, some remarks on economic impact analysis as a policy evaluation tool are made. Then, the organization and contents of small industry programmes in Indonesia are described in short. This is followed by a description of the field survey, the data, specifications and methodology used, a discussion of the results and some concluding remarks.