When chest symptoms recur in a patient who underwent percutaneous transluminal coronary angioplasty (PTCA), it is necessary to rule out restenosis (R). Three main noninvasive tests suggest the presence of R: exercise stress test (XT), myocardial perfusion imaging (MPI) and stress echocardiography (s-echo). The objectives of this review were: (1) to estimate the pretest probability of R as a function of time after PTCA in symptomatic patients and (2) to obtain an approximation of the diagnostic parameters of the XT, MPI and s-echo for detecting R. A MEDLINE search (English-language, years: 1980-2001) was conducted to identify studies examining post-PTCA functional testing for diagnosing R. Data from the studies were pooled. Comparing studies was often difficult due to varying methodology in the studies. Pretest probability of R in symptomatic patients increases in a nonlinear fashion from 20% or less at 1 month, to nearly 90% at 1-year postangioplasty. The approximated accuracy of the XT, MPI, and s-echo for detecting R was 62, 82 and 84%, respectively. During the first month after PTCA, none of the noninvasive modalities is able to accurately detect R. Late (7-9 months) after PTCA, the pretest probability of R is high and therefore the noninvasive measure may be spared. Our analysis suggests that MPI and s-echo should be preferred over the XT for diagnosing R.