Research tournaments are predicted to encourage R&D. Recent theoretical dev
elopments in research tournaments are grounded in search theory. Although t
he economic intuition behind tournaments is straightforward, computing equi
librium strategies is complex. The participants compute a stopping rule bas
ed on the number of participants, the prize and the cost of research. It is
an empirical question whether agents will behave as predicted or will empl
oy simple "rule of thumb" strategies such as faking a predetermined number
of draws. This paper reports the results of a series of laboratory experime
nts designed to test the predictions of the search model of tournaments. (J
EL C9, D8).