Summary Rhododendrons ( Rhododendron ponticum L. and Rhododendron luteum Sweet) dominate the understories of the mesic forests of the Black Sea Region (BSR) of Turkey. They dramatically reduce forest growth and regeneration and local plant diversity. This paper reports the results of a large rhododendron control experiment established on two different sites in the western and eastern BSR of Turkey 5 years after treatments (YAT). The paper also presents the second-year results of a second experiment in which the foliar herbicides of the previous experiment were tested on rhododendron at much lower rates on a western BSR site. Five YAT, grubbing and foliar spraying were still the best rhododendron control methods in the fi rst experiment. Cut-stump spraying provided an intermediate level of woody control. Hand-cutting was ineffective on R. ponticum and did not signifi cantly differ from the control treatment in rhododendron basal area 5 YAT. The performance ratings of foliar triclopyr ester and imazapyr for woody control in both experiments were rate dependent. At high rates ranging between 2.6 and 5.8 kg ae ha − 1 , foliar imazapyr controlled both rhododendron species signifi cantly better than foliar triclopyr ester in the fi rst experiment, suggesting enhanced imazapyr translocation to the roots. Some off-target damage was observed in the beech overstory for imazapyr at high rates. The performance ratings of these foliar herbicides were signifi cantly reversed in the second experiment 2 YAT, where much lower rates were used (0.3 – 2.0 kg ae ha − 1 ) than in the fi rst experiment. Insuffi cient imazapyr accumulation in rhododendron roots might account for the poor impact of this herbicide. No beech damage was apparent from any of the herbicides in the second experiment. The low-rate foliar triclopyr ester is recommended for effective and cost-effi cient rhododendron control.