This paper reassesses the UK results of significant abnormal returns f
rom directors' trading for a new sample of directors' trades 1984-1986
, and finds that abnormal returns tend to be concentrated in smaller f
irms. When an appropriate benchmark portfolio is used, it is found tha
t the significance of the abnormal returns is substantially reduced, w
ith the implication that directors' trading does not yield particularl
y high profits to either the directors themselves or to an outside inv
estor mimicking those trades.