Ten timber beams were tested, including three reference normal beams, three unstrengthened lattice beams with different configurations, and four lattice beams strengthened by different methods. The test results showed that the ultimate flexural strength and initial flexural stiffness of unstrengthened lattice timber beams are lower than the reference normal beams. Compared with the unstrengthened lattice beam, the ultimate flexural strength, initial flexural stiffness and mid span displacement at the ultimate state increase by 54%, 34% and 63%, respectively, for the beam bonded with a CFRP sheet to the tension face, 30%, 195% and -54%, respectively, for beams with their spaces at the middle filled with structural adhesive or cement grouting, and 92%, 251% and -48%, respectively, for the beam strengthened with both structural adhesive filling and a CFRP sheet bonded to its tension face. The strain distributions in the upper and lower legs are nearly linear through their respective depth at middle span during the whole loading process for both unstrengthened lattice and CFRP strengthened lattice beams but the plane section assumption is not satisfied as a whole. For beams strengthened with structural adhesive or cement grouting fillings, the strain distributions along the entire height at middle span are nearly linear during the whole loading process so a plan section remains almost plan. This study showed that external bonding of CFRP sheets is an effective method for enhancing the ultimate strength and filling with structural adhesive is an effective way for increasing the initial flexural stiffness. The finite element analysis results of lattice timber beams can completely satisfy the engineering precision requirement.
