With the judicious selection of triplet energy donor (D) and acceptor (A) pairs, a laser flash photolysis procedure has provided a sensitive method for the study of triplet energy transfer in rigid polymer films. By monitoring changes in triplet-triplet (T-T) absorptions the kinetics of triplet energy transfer were evaluated at short time scales, and overall energy-transfer quantum yields were also obtained. Combinations of xanthone- or thioxanthone-type donors and polyphenyl acceptors were particularly suited to these measurements because the former have high intersystem-crossing quantum yields and the latter have very high extinction coefficients for T-T absorption. For exothermic transfer most of the energy transfer that occurred within the lifetime of triplet D (D-3) took place in less than a few microseconds after D-3 formation in poly(methyl methacrylate), and triplet A yields were limited largely by the number of A molecules in near contact with D-3. The kinetics of triplet energy transfer were modeled using a modified Perrin-type statistical arrangement of D/A separations with allowance for excluded volume in combination with a Dexter-type formula for the distance-dependent exchange energy-transfer rate constant. Experimental observations were best explained by constraining D/A separations to reflect the dimensions of intervening molecules of the medium. Rate constants, k(0), for exothermic energy transfer from ID to A molecules in physical contact are approximately 10(11) s(-1) and very similar to triplet energy-transfer rate constants determined from solution encounters. Energy-transfer rate constants, k(r), fall off as approximately exp(-2r/0.85), where r is the separation distance between D and A centers in anastroms. Exchange energy transfer is not restricted to ID and A in physical contact, but at <= 0.4 M A at least 85% of the energy transfer arises from interaction of D-3 with a single nearest-neighbor A molecule. The modified Perrin model was also applied to quantum yields of quenching in rigid media. Comparison to the simple Perrin model for quenching shows that the latter may be adequate as long as molecular volumes are accommodated in the Perrin expression. Under these conditions the critical radius, r(c) corresponds to the D-3/A separation at which the effective rate constant for energy transfer equals the inverse of the D-3 lifetime.