We examined the relationship between firm births and job creation in Great Britain. We used a new data set for 60 British regions, covering the whole of Great Britain, between 1980 and 1998. The relationship between new firm start-ups and employment growth has previously been examined either with no time lag or with only a short period lag. We found, for Great Britain as a whole, no significant relationship between start-ups and employment creation in the 1980s, but a negative relationship for the 'low enterprise' area of Northern England. For the 1990s, we found a significant positive relationship for Great Britain as a whole but for Scotland, which focussed policy on start-ups, a negative relationship. We feel this raises questions over policies designed to raise rates of new firm formation as a strategy for employment creation, particularly in 'low enterprise' areas. This paper was previously published as Research Report H200108 and as Scales Paper N200202. The major change compared to the previous version entails separate studies of region-specific deviations in the relationships. The effect of the number of start-ups on growth was found to be significantly smaller in regions with relatively low numbers of start-ups, also called the 'Upas Tree effect'.